Aortic stenosis (AS) is a common, progressive valvular lesion with a poor prognosis when left untreated (1). Octogenarian patients with AS have a high prevalence of coexisting conditions and may be at increased risk of periprocedural morbidity and mortality when undergoing surgery (2).

Surgical aortic valve replacement (SAVR), first performed in the 1960s, became the gold standard for treating AS (3), but surgeons were initially reluctant to operate on older patients (2). Minimally invasive SAVR, a modification of the original technique, was developed with the purpose of minimizing operative trauma and reducing postoperative mortality and morbidity (4). Balloon valvotomy was the first procedure to emerge as a possible endovascular treatment (3). Aspiration, an even less-invasive method—suitable for patients with little or no chance of surviving surgery—led to implantation of the first transcatheter aortic valve implantation (TAVI), in 2002. Since then, TAVI for AS has been increasingly adopted in clinical practice (5).

With well-established surgical methods it was not easy to prove the non-inferiority of TAVI (6). However, with meticulously designed studies conducted over the past decade, advocates of TAVI not only demonstrated its non-inferiority in high-risk patients, they proved its benefits, and expanded its indication for use in intermediate- and low-risk patients (7–11).

Advances in treatment techniques have led to different options to consider for our patients. The first mention of TAVI was in the 2008 guidelines on the treatment of AS. More recently, evidence from clinical trials has now positioned TAVI as a suitable treatment option for AS (12–14).

The objective of this study was to compare clinical outcomes among octogenarian patients undergoing TAVI or SAVR treated at our facility during the past decade.

The results of this single-center, retrospective cohort study showed that the type of procedural approach for AS—whether TAVI or SAVR—did not affect in-hospital, 30-day, or 1-year mortality. Therefore, the anticipated advantages of using a less invasive technique did not appear to influence early or mid-term outcome. Rates of PPI and paravalvular leak were more frequent in the TAVI group, whereas new onset AF and acute kidney injury were more common after SAVR. After stratification by STS score, there were no differences in survival between groups, whereas a statistically significant difference was found in low-risk and intermediate-risk groups in terms of moderate or severe paravalvular leak, in favor of SAVR, and in acute kidney injury and new-onset AF, in favor of TAVI.

With the development of TAVI over the past two decades and its introduction into routine clinical practice, it was inevitable to ask whether TAVI is the best treatment for all octogenarians (16, 17). The results of the Placement of Aortic Transcatheter Valves (PARTNER)-1 trial—the first prospective randomized study comparing TAVI, best medical therapy, and surgical therapy in high-risk patients—were published in 2011. Rates of all-cause death at 1-year were similar in the TAVI and SAVR groups, and were superior to best medical therapy (7). PARTNER 2A and Surgical Replacement and Transcatheter Aortic Valve Implantation (SURTAVI), both of which were designed to compare outcomes in intermediate-risk patients, also displayed non-significant differences between TAVI and SAVR in the 1-year non-hierarchical composite of all-cause death or disabling stroke (8, 9). Recent trials focusing on low-risk patients—PARTNER 3 and Medtronic Evolut Transcatheter Aortic Valve Replacement in Low Risk Patients (EVOLUT LRT)—opened new doors for TAVI and established the role of this approach in all risk groups. However, caution is needed when applying the findings from RCTs, which are subject to strict enrollment criteria, to the spectrum of patients with AS treated in routine clinical practice (18).

Our primary outcome, 1-year survival, was not influenced by the type of procedure. Recent studies focusing solely on octogenarians are scarce and show similar outcomes for death (17). Other reports use different composite outcomes, making direct comparisons difficult (19–22). A meta-analysis by Witberg et al. (23) suggested a trend toward reduced rates of 1-year death with TAVI in RCTs in low-risk populations, whereas observational studies with PSM in the same meta-analysis showed a trend toward increased rates of death with TAVI. In our study, there was no statistically significant difference in hospital or 30-day mortality in the entire cohort or in the low-risk or intermediate-risk groups between TAVI and SAVR. Comparisons in the high-risk group were not possible because of the small number of patients who underwent SAVR. After performing PSM in our population there was a statistically significant difference between the groups in terms of hospital stay, in hospital and 30-day mortality. Mollmann et al. (24) reported lower in-hospital mortality with TAVI in an all-comer AS population treated in Germany. However, a systematic review and meta-analysis by Moss et al. (17) in octogenarians showed no differences in 30-day mortality between TAVI and SAVR.

The rate of PPI in the present study was higher in the TAVI group, in which 15.0% of patients developed postprocedural conduction disturbances. The 9.3% of patients requiring PPI after SAVR is similar to the rate reported in other studies in octogenarians (17, 25) as well as in major TAVI studies (7–11). Interestingly, even though more than half of our patients in the surgical arm had a sutureless self-expanding (Perceval or Enable) or balloon expandable (Intuity) valve implanted, the incidence of PPI was low (26–28). Data regarding the effect of new PPI after TAVI on mortality and morbidity are inconclusive (29). When expanding the indication for TAVI to the younger population, it is important to consider that PPI-related complications are more frequent in younger patients (30).

Postoperative moderate-to-severe paravalvular leak was present in approximately 10% of TAVI patients and in only 0.8% of SAVR patients. Compared to previous RCTs (7–10) our TAVI population presented with relatively high incidence of postoperative moderate-severe paravalvular leak. This could in part be attributed to the fact that we only had access to early echocardiographic findings, with the ultrasound of the heart performed during hospitalization or before the discharge. At least a 30 day follow-up ultrasound would be necessary to properly evaluate our cohort. The PARTNER trial showed that increased severity of paravalvular leak is associated with higher 2-year mortality (31). Hagar et al. (32), however, found no association between paravalvular leak and 1-year mortality in a slightly younger population (mean age 74 years).

Neurological complications in our study were in line with the PARTNER-2 and SURTAVI results for the SAVR group (3% for cerebrovascular stroke or TIA), whereas results in the TAVI group were similar to those of the PARTNER 3 and EVOLUT LRT results, with a much lower incidence (0.7% of cerebrovascular stroke or TIA) despite a higher incidence of patient-related risk factors (33). It is difficult to explain this difference between TAVI and SAVR groups. We could attribute the incidence of neurological complications to unmeasured patient and periprocedural characteristics or operator skill and experience.

New onset AF (within 30 days of the procedure) was significantly more frequent in the SAVR group and is similar to the rate reported in RCTs (7–11) and retrospective studies (34–36). It is important to consider that almost half of the patients in the TAVI group had AF preoperatively. A number of studies evaluated possible risk factors and preventive measures. Older age and moderate-to-severe left atrial enlargement were the only two consistent independent factors of prolonged postoperative AF (34). Mathew et al. (37) showed that each consecutive decade was associated with a 75% increase in the risk of developing postoperative AF. Postoperative pericardial effusion is frequently observed after surgery, and experimental studies indicate that even a small amount of effusion may trigger AF due to mechanical compression of the atria, local inflammation, and oxidative stress (38).

Acute kidney injury was present in 32.7% of our SAVR patients. In studies conducted before the use of common diagnostic criteria [RIFLE (Risk, Injury, Failure, Loss of kidney function, and End-stage kidney disease) or AKIN (Acute Kidney Injury Network)], the reported incidence of acute kidney injury was lower than in our study (39, 40), but later studies showed similar findings, with acute kidney injury present in approximately one-third of patients after SAVR (41–44). An important finding was that even small changes in serum creatinine concentration are associated with adverse outcomes (45). In SAVR, the specific reason is cardiopulmonary bypass that causes inflammation and hemodilution and is accompanied by periods of low pressure and flow rates. The duration of cardiopulmonary bypass and aortic cross-clamp time were linked to the development of postoperative acute kidney injury (46). In our study, the average aortic cross clamp time was 47.0 ± 17.7 min, with a 50% higher occurrence of acute kidney injury in the SAVR group with an aortic cross-clamp time exceeding 50 min. Contrast exposure, embolization, hemodynamic instability, and access route play important roles in acute kidney injury after TAVI. Transapical access, compared with transfemoral access, was shown to be an independent predictor of acute kidney injury following TAVI (45), which was also seen in our study.

Major or life-threatening bleeding occurred in 11.0% of TAVI patients and in 6.7% of surgical patients, but the difference was not statistically significant. Octogenarians are a more vulnerable population because of their advanced age, comorbidities and frailty, independent of the STS score (47, 48), and a higher rate of bleeding complications was expected. The low rate of bleeding in SAVR group could be a consequence of using minimally invasive techniques (76.6% of patients underwent minimally invasive SAVR), and the high rate of bleeding complications in the TAVI group could be due to the large size of the delivery systems used.

Stratification by STS score into three groups showed that quite a sizeable number of patients in TAVI group were low risk. With rapid expansion of TAVI volumes, the risk profile of patients treated in everyday practice has been lower than those included in RCTs, which are the basis for practice guidelines (23). Our results showed no statistically significant differences between specific risk groups (TAVI vs. SAVR) in 30-day or 1-year mortality. Further, the stratification showed the expected significant difference between TAVI and SAVR in low- and intermediate-risk groups in occurrence of acute kidney injury, paravalvular leak, and new onset AF. Analysis of the high-risk group was not possible because it included only 2 SAVR patients.

The promising results from TAVI trials have fundamentally changed the way we treat octogenarians. It is important to take into consideration that not all patients, due to their advanced age, are automatically high risk (49). STS-PROM and STS/American College of Cardiology in-hospital mortality scores are superior to EuroSCORE I, EuroSCORE II, and the German AV Score (50). Despite more than 10 years of clinical experience with TAVI, a reliable risk score model that includes frailty is not yet available. Certain octogenarians with a low STS score may benefit more from a surgical procedure and avoid the possibility of paravalvular leak, which affects overall survival. A reliable risk score would help us recognize such patients and tailor their treatment individually. Treatment options should remain open, and regular revision of “real-life” results should ensure that we can offer patients the best treatment option. Detailed explanation of possible complications and balanced information regarding outcome after each procedure should be given to obtain informed patient consent prior to the choice of procedure.

This retrospective study involving an octogenarian “real-life” population treated for isolated severe AS provides insights into our decision-making and shows that similar short- and mid-term results were obtained with both TAVI and SAVR in low- and intermediate-risk (STS score) groups. We showed that the type of procedural approach did not significantly affect 1-year mortality. The expected advantage of a less invasive technique did not influence early outcome, with similar rates of in-hospital mortality and length of hospital stay in both groups. PPI and paravalvular leak were more frequent in the TAVI group, whereas new onset AF and acute kidney injury were more common in the SAVR group. In light of our findings, the treatment strategy in each institution should be adopted according to its own outcome data and facilities.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the Ethical committee of Republic of Slovenia. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

TK and MB: conception and design of the study. TK and AK: data collection. TK, NL, AK, DŠ, ZF, and MB: analysis and interpretation. TK, NL, DŠ, ZF, and MB: drafting and critical review of the article. All authors have read and approved the final manuscript.