Multimodal protocol influence on stroke and neurocognitive deficit prevention after ascending/arch aortic operations

Various techniques are used for brain protection during aortic surgery. Rather than evaluate each factor separately, we evaluated the early outcome of a multimodal protocol (mannitol, thiopental, MgSO4, lidocaine, CO2 field flooding, Leukoguard filter, head ice packing, electroencephalographic arrest at 20 degrees C, alpha-stat, increasing right subclavian artery cannulation, and antegrade/retrograde brain perfusion) for brain protection.

Prospectively collected data were analyzed on 403 ascending or arch aortic operations including 199 (49%) arch replacements conducted between July 25, 1991, and September 25, 2001. The mean age was 61.6 years (range 22 to 91 years); 48 (12%) had Marfan syndrome; 141 (35%) had dissection; 134 (33%) had composite grafts inserted; and 138 (34%) had concurrent coronary bypasses performed.

Stroke occurred in 2.0% (8/403) (3 permanent, 5 transient), clinical neurocognitive deficits in 2.5% (10/403) either by testing or patient complaint 2 to 3 weeks after surgery, and 98% (395/403) were 30-day survivors. Univariate predictors of stroke, neurocognitive decline, or death were the following: for stroke, aorta symptom severity grade (1 to 4) (p = 0.001), pump time (p = 0.001), arrest time (p = 0.001), macroscopic atheroma (p = 0.041), concurrent descending/thoracoabdominal aneurysm (p = 0.036), and highest blood rewarming temperature (p = 0.043); for neurocognitive decline, degree of cooling (p = 0.046), pump time (p = 0.001), cooling time (p = 0.001), day extubated (p = 0.042), and antegrade brain perfusion (p = 0.004); for death, pump time (p = 0.001) and clamp time (p = 0.011). The multivariable independent predictors of stroke, neurocognitive decline, or death were the following: for stroke, aorta symptoms grade (p = 0.025), peripheral vascular disease (p = 0.043), and pump time (p = 0.015); neurocognitive decline, preoperative New York Heart Association dyspnea class (p = 0.022), pump time (p = 0.05), arrest time (p = 0.06), day extubated (p = 0.042), and antegrade perfusion (p = 0.023); and for death, pump time (p = 0.018).

Pump time continues to be the most important predictor of adverse events. The benefit of antegrade or retrograde perfusion remains unproven, partly because of the low event rate (< 2.5%) but may be beneficial for prolonged circulatory arrest. Embolic material either from macroscopic atheroma, descending or thoracoabdominal aneurysms, or associated with peripheral vascular disease, increases the risk of stroke. Preoperative symptoms influence outcome.