New analysis confirms that the number, size and total volume of acute brain infarctions are each associated with clinical ischemic strokes, disabling strokes and worse recovery in patients undergoing transcatheter aortic valve replacement (TAVR). Findings also reveal that 4-6 lesions, defined by diffusion-weighted magnetic resonance imaging (DW-MRI), maximum individual lesion volume (ILV) with a threshold of 216 mm3 and total lesion volume (TLV) with a threshold of 440-550 mm3 has “excellent” discrimination in identifying ischemic and disabling stroke. “A clinically meaningful TLV threshold of ≥450 to 550 mm3 represents a major embolic risk after TAVR that occurred in approximately 30% of patients in our study,” said the paper’s authors. “[This] accounted for 76% of all strokes and 80% of all disabling strokes. “Although it is unreasonable to think that any surrogate can perfectly predict stroke, we show that dichotomizing TLV at a threshold ≥500 mm3 is remarkably effective at identifying very low and very high risk of stroke such that cerebral embolic protection (CEP) could use this threshold in a hierarchical analysis to look for a potential benefit.” Patient-level pooled analysis Results of the patient-level pooled analysis, which was published in the Aug 20 issue of the Journal of the American College of Cardiology, revealed that out of the 495 patients undergoing TAVR and completed the predischarge DW-MRI, the rate of clinical ischemic stroke was 6.9% at 30 days. Acute ischemic brain injury was seen in 85% of patients with 5.5±7.3 discrete lesions per patient and with a mean ILV of 78.2±257.1 mm3, and mean TLV of 555±1,039 mm3. The C-statistic was 0.84 for TLV, 0.81 for number of lesions, and 0.82 for maximum ILV in predicting ischemic stroke. On the basis of the TLV cutpoint as defined by receiver operating characteristic (ROC), patients with a TLV >500 mm3 (vs TLV ≤500 mm3) had more ischemic stroke (18.2% vs 2.3%; P<0.0001), more disabling strokes (8.8% vs 0.9%; P<0.0001) and less complete stroke recovery (44% vs 62.5%; P=0.001) at 30 days. “Our study (also) adds to our mechanistic understanding of stroke during TAVR, whereby larger showers of emboli (lesion counts were >3.5-fold higher in stroke patients) contributed more than larger individual lesions to the overall >5.5-fold increase in TLV associated with stroke,” said the paper’s authors, led by Alexandra J. Lansky, MD, from Yale School of Medicine in New Haven, Connecticut. “Although risk factors for identifying patients at high risk of embolic stroke after TAVR are challenging because of the low frequency of stroke events, our observation supports measures to limit the total number and ultimately the total volume of brain injury.” Key takeaways Commenting on the analysis, which was also published Monday online, Samir Kapadia, MD, and Amar Krishnaswamy, MD, from the Cleveland Clinic, and Michael Mack, MD, from Baylor Scott & White Health in Dallas, Texas, highlighted key takeaways arising from this analysis. This included the data obtained in the reported clinical trials with imaging endpoints, where approximately 18% of patients were not able to undergo magnetic resonance imaging (MRI). According to the experts, this could be an important consideration when designing studies with primary imaging endpoints. “This rate is similar to what was encountered in the Cerebral Protection in Transcatheter Aortic Valve Replacement (SENTINEL) trial (21.2%),” they said. “Frequently, the reasons for not having an MRI are clinical events that can potentially confound the interpretation of results (e.g., hemodynamic instability or other complications of the procedure).” The experts go on to point out that stroke ascertainment is influenced by imaging, leading to a much higher ascertainment rate of clinical events (6.9% in this analysis), which was similarly high (6.8%) in the SENTINEL trial. “This observation raises an important issue when interpreting clinical trials with lower stroke rates with only clinical evaluation without routine imaging,” the experts said. “Therefore, the clinical implications of imaging assisted stroke ascertainment require further studies to truly understand the short- and long-term ramifications.” What data is required? The accompanying editorial comment then posed the question of what additional data was needed for the use of cerebral embolic protection devices (CEPD) during TAVR. Kapadia, Krishnaswamy and Mack commented that while there were several indications that CEP reduced clinical strokes, improved outcomes of strokes when they happened and was safe to use, the evidence from large randomized controlled clinical trials has not been definitive and therefore device use had been limited. “Is it necessary to get this conclusive evidence?” they asked. “This is certainly needed for reimbursement considerations and to convince skeptics who appropriately question the nonrandomized data.” They went on to focus on the patient population that benefited the most from CEPD, which currently remained unclear and reiterated the need for additional data to identify the clinical and/or anatomic characteristics of the population. The experts also commented on the limitations restricting the use of MRI-related endpoints for clinical trials. “Excluding nearly 20% of endpoints could certainly be problematic for robustly assessing a trial’s outcome if that outcome measure is primarily MRI based,” they said, in conclusion. “Further, protocols for MRI acquisition and interpretation can also be quite variable and require standardization. Finally, the expense of brain MRI can be a limiting factor for clinical trials.” Study approach Patient-level data were pooled from 4 prospective TAVR embolic protection studies, with consistent predischarge DW-MRI acquisition and core laboratory analysis. A total of 495 of 603 patients undergoing TAVR completed the predischarge DW-MRI. These patients had a mean age of 80.6±7.3 years, of which 55.8% were male. C-statistic was used to determine the best DW-MRI measure associated with clinical stroke. The primary endpoints for this analysis were ischemic stroke, including ischemic stroke with hemorrhagic conversion and disabling ischemic stroke at 30 days. Sources: Lansky AJ, Grubman D, Dwyer III MG, et al. Clinical Significance of Diffusion-Weighted Brain MRI Lesions After TAVR Results of a Patient-Level Pooled Analysis. J Am Coll Cardiol. 2024;84:712–722. Kapadia S, Krishnaswamy A, Mack M. Acute Brain Infarctions and Periprocedural Stroke Implications for Evaluating Cerebral Embolic Protection Devices. J Am Coll Cardiol. 2024;84:723–725. Image Credit: New Africa – stock.adobe.com