Serial computed tomography angiography-based computational fluid dynamics characterizes reduced near-renal tear exchange in kidneys that develop atrophy after aortic dissection repair

Background Renal atrophy may develop after aortic dissection repair despite an anatomically patent renal artery on computed tomography angiography (CTA). This exploratory study compared serial near-renal aortic hemodynamics between kidneys with and without later renal-volume loss. Methods Six index kidneys from six men were studied: three with ≥30% loss of reference volume (atrophy group) and three age-matched comparators. Patient-specific computational fluid dynamics (CFD) models were built from preoperative, early postoperative and late follow-up CTA; within each patient, inlet and outlet (three-element Windkessel) boundary conditions were identical across time points. All 90 true-lumen/false-lumen (TL/FL) communications were characterized by distance from the renal-ostial plane, direction, pressure difference and trans-tear flow. TL/FL relationships were quantified at the renal-ostial plane, and bilateral renal-artery flow was computed directly. Renal volume was measured by three-dimensional CTA segmentation. Results All six index kidneys were supplied from the FL before repair. Index-kidney volume changed by −53.0%, −64.1% and −50.6% in the atrophy group and by −13.8%, +0.8% and −10.7% in comparators. No hemodynamic measure separated the groups before repair. From the early postoperative model onward (3–37 days), two aortic-level flow measures were lower in every atrophy kidney than in every comparator: the FL flow-magnitude fraction at the renal-ostial plane (0.03–0.10 versus 0.13–0.79) and total peak-systolic trans-communication flow (37.8–45.1 versus 61.8–185.3 mL/s). Index renal-artery flow separated the groups only at late follow-up (−29.4%, −32.6% and −63.9% versus +5.9%, −7.4% and −3.0%), as did tear-exchange flow directed toward the renal-ostial level (1.9–19.1 versus 38.7–130.9 mL/s). FL/TL pressure ratios did not separate the groups at any time point. In one patient-derived geometry, sequential virtual closure of idealized communications redistributed and reversed flow in the remaining communications; index renal-artery flow after the final closure was similar to baseline. Conclusion In kidneys that later atrophied, the FL flow contribution at the renal level was lower than in comparators early after repair, before the modeled decrease in renal-artery flow and, in two of three kidneys, before substantial volume loss. Given three kidneys per group and multiple measures, these associations are hypothesis-generating and do not establish causation, renal parenchymal perfusion or a treatment threshold.