Long Term Endoleak Outcomes After Thoracic Endovascular Aortic Repair: A 25 Year Single Centre Experience

Introduction: Since its introduction in 1994, thoracic endovascular aneurysm repair (TEVAR) has become standard practice in the management of a broad spectrum of descending thoracic aorta diseases. Despite several publications focusing on the short and mid-term outcomes of TEVAR, few analyses encompassing a multidecade time span are available. While the efficacy and peri-operative morbidity and mortality of this technique have been extensively investigated, data regarding long term survival and the endoleak rate are still limited. Methods: This was a retrospective, observational, single centre cohort study of TEVAR patients between November 1995 and December 2020 (local ethics committee approval no. 507/2018). Missing mortality information (dates and causes) and eventual re-interventions performed at other centres were retrieved by the Department of Public Health. The primary endpoint of this study was freedom from endoleak in the four stent graft generations during the follow up period. No re-interventions for type II endoleak were registered; they were therefore excluded from the cumulative endoleak rate. The first generation (GEN1) included: Vanguard (Boston Scientific); AneuRx and Talent (Medtronic); Stentor (Mm Tec); Excluder (WL Gore); Endologix (Endologix); and EndoFit (LeMaitre). The second-generation (GEN2) included TAG (WL Gore) and TX (Cook Medical). The third (GEN3) included Relay Plus (Bolton Medical); Valiant Captivia (Medtronic); and Zenith Alpha and Custom-made (Cook Medical). The fourth (GEN4) included Relay Pro and custom made (Bolton Medical); Conformable C-TAG (WL Gore); Navion (Medtronic); Standard and custom made thoraco-abdominal devices (Cook Medical); Nexus (Endospan); E-Vita, Colt, and custom made thoraco-abdominal devices (Jotec); and Najuta (Kawasumi). Aorta related survival was considered whenever the cause of death was ascribable to the treated aortic segment. Results: A total of 509 stent grafts deployed into the thoracic aorta were included with a mean ± SD follow up of 44.3 ± 42.5 months. Early and late results are depicted in Table 2. Freedom from endoleak was 84.6%, 70.4%, 56.0%, and 51.5% at one, five, 10, and 15 years, respectively (Fig. 1). Freedom from endoleak at one year was 90.4%, 96.2%, 86.4%, and 96.6% for GEN1, GEN2, GEN3, and GEN4, respectively. Freedom from endoleak at five years was 65.6%, 61.4%, 76.2%, and 69.1% for GEN1, GEN2, GEN3, and GEN4, respectively (p = .37). The first two generations demonstrated a higher endoleak rate, compared to the two most recent generations (27.2 vs. 18.2%; p = .043). The proximal landing in zone 1 (odd ratio [OR] 2.2; p = .050) and GEN1 (OR 2.0; p = .014) was an independent risk factor for endoleak. Proximal landing in zone 5 (OR 0.2; p = .055) and GEN4 (OR 0.3; p = .001) was an independent protective factor for endoleak. GEN1 was found to be a risk factor for type Ia endoleak (OR 2.1; p = .059). Type Ib endoleak demonstrated a negative association with distal landing in zone 5 (OR 4.9; p = .001). Conversely, a positive association was highlighted with distal landing in zone 4 (OR 0.2; p = .036) and GEN4 (OR 0.1; p = .012). Overall survival was 75.3%, 44.4%, 27.2%, and 17.6% at one, five, 10, and 15 years, respectively. Survival distinguished as non-related versus aortic related was 86.7% versus 23.5%, 52.7% versus 9.8%, 32.9% versus 2.0%, and 21.2% versus 0% at one, five, 10, and 15 years, respectively (p < .000; Fig. 2). Conclusion: Potentially, half of patients could develop an endoleak at 15 years, highlighting the importance of life long radiological follow up after thoracic stent grafting. Newer stent graft generation demonstrated better long term endoleak results, despite an increasing indication of more challenging aortic morphology, as well as to extensive pathology (arch and/or thoraco-abdominal), during the 25 years analysed. This real world TEVAR study over a considerable time span yielded the conclusion that the older generation should be followed up strictly. Likewise, landing in the most extreme zones (e.g., proximally in zone 1 and distally in zone 5) required special attention owing to their higher tendency to leak. The efficacy of TEVAR was confirmed in preventing long term aortic mortality, with most deaths occurring within 30 days.