Operation in the gray zone: is SAVR still useful in patients aged between 75 and 80 years?

ABSTRACT

Introduction

Aortic valve replacement is the only effective treatment for symptomatic aortic valve disease. Transcatheter implantation has been introduced as alternative for surgery (SAVR), but its penetration varied widely. For high-risk octogenarian patients, the advantages of TAVI are clear. Patients between 75 and 79 years of age could be considered as “the gray zone.”

Methods

We compared the outcome of SAVR with or without concomitant procedures between patients younger than 75 versus 75–79 years for their preoperative profile, operative parameters, and postoperative outcomes.

Results

Older patients had a higher risk score with more cardiovascular comorbidity and renal dysfunction. They required more concomitant CABG but less mitral valve repair. Cardiopulmonary bypass time was similar. In-hospital complication rates, need for blood transfusion, and length of stay in intensive care unit were higher in older patients. Thirty-day mortality was not significantly increased, Euroscore II overestimated mortality in both age groups. Age over 75 years was not an independent predictor for mortality. Need for emergent SAVR was only predictive in the older group. Long-term survival was acceptable. Age over 75 years was not the dominant predictor.

Conclusions

SAVR remains a valid option to treat symptomatic aortic valve disease in the age group of 75–79.

Plain Language Summary

Symptomatic aortic valve disease is a highly lethal condition, and its only effective treatment is replacement of the diseased valve. Surgery is the oldest mode of treatment but carries a risk in older patients, although surgery has been performed with success in patients of 80 years and older. For these patient groups, the implantation of a valve prosthesis through a transcatheter approach is a less invasive alternative, while for younger and low-risk patients, surgery remains the most important option. Patients older than 75, but younger than 80 years could be considered as the “gray zone.” For this reason, we compared patients in this age category with patients younger than 75 years with respect to their preoperative profile (risk score and concomitant cardiac and non-cardiac diseases), operative data and postoperative outcomes (need for postoperative resources, postoperative complications, and mortality in the hospital and at long-term). We found that older patients had more concomitant heart conditions and worse kidney function. Their risk (Euro II) score was also higher. They also needed more coronary artery surgery. The older patient group suffered more complications but had no higher need for resources, except for blood transfusion. They also stayed one day longer within the intensive care unit. Hospital mortality in the older group was not significantly increased. The risk score overestimated mortality in both age groups. Long-term survival was favorable in both groups. The need for emergent surgery affected short- and long-term mortality only in the older age group.

1. Introduction

Compared to some malignancies, severe aortic valve stenosis has a poor prognosis if the valve is not replaced [1]. Surgical aortic valve replacement (SAVR) is the oldest mode of treatment but more recently, transcatheter aortic valve implantation (TAVI) has been introduced as an alternative treatment [2]. TAVI has profoundly changed the management of patients with degenerative aortic valve disease. The use of this procedure also seemed beneficial in patients with intermediate and low-risk [3]. European and American guidelines favored the use of TAVI in patients of 75 years or older and with an Euroscore of over 8%, but also in patients with a history of prior cardiac surgery, severe frailty, sequelae of prior chest irradiation, a porcelain aorta, and a chest deformation. The use of SAVR was favored in patients with an age below 75 years and with an Euroscore of less than 4, a history of endocarditis, a bicuspid aortic valve, a thrombus within the left ventricle, a risk for coronary obstruction with TAVI, presence of multivessel disease which requires revascularization, or of an aneurysm of the ascending aorta [4,5]. However, the penetration of TAVI varied widely [6], with long waiting lists in some areas as a consequence [1]. One nationwide study showed that the number of aortic valve replacement of all types has doubled in a time span of 10 years. This showed that TAVI had become the dominant form of aortic valve replacement in 2018, while the number of SAVR had decreased. However, the effect of the introduction of TAVI on the preoperative patient profile and volumes of SAVR in different regions were not uniform across counties and regions [6]. The use of TAVI in Belgium was restricted from its introduction [7,8]. This policy differed from surrounding larger European countries and seemed mainly to be driven by the cost of the valve prosthesis. A downward trend for the costs for TAVI has been observed between 2012 and 2021 in the US and Canada. This could be related to the advancements in medical technology and also could allow its more widespread use. TAVI, however, is still more expensive compared to SAVR and depends on individual patient characteristics, access site and choice of the valve prosthesis. Medicare reimbursed the majority of the procedures in the US [9]. TAVI might become more cost-effective than SAVR in the future, but the various comorbid conditions should be taken into account. Economic as well as procedural issues such as the labeling of anatomical inoperability and surgical risk were also driving the reimbursement policy for TAVI in Belgium. Comorbid conditions such as pulmonary artery hypertension, but also the need for postprocedural pacemaker implants and the more difficult valve-in-valve procedure contributed to this policy [7]. For these reasons, TAVI reimbursement has been restricted: a maximum number per year and the number per center was linked to the number of SAVR [7]. This improved partially after 2017. The main drivers were the price of the valve and length of stay. A TAVI procedure with a SAPIEN 3 device provided meaningful clinical and cost benefits over SAVR. The initial procedure costs were higher for TAVI, but its advantages were related to quality-adjusted life years and costs reductions over time. These findings applied to rehabilitation, disabling stroke, treated atrial fibrillation, and rehospitalization [8]. Another Belgian retrospective single-center analysis, comparing sutureless SAVR with TAVI, showed that length-of-stay was significantly longer after SAVR and costs were higher if the valve prosthesis was excluded from the calculation. These costs were mostly driven by the stay on ICU and on the ward. After inclusion of the prosthesis, SAVR was shown to be cost-saving compared to TAVI [10]. With the introduction of rapid deployment valves, shorter operation times and the possibility for minimally invasive access, SAVR [11] could also be applicable for high risk patients of 75–79 years and even in selected octogenarian patients [12,13]. Because of the paucity of data about the results of SAVR in the age class 75–79 years, which could be considered as “the gray zone,” following research questions were addressed for patients in this age class versus their younger counterparts

–

1. In what do these groups differ in preoperative and operative characteristics?

2. In what do these groups differ in 30-day postoperative outcome and long-term survival?

3. Can age over 75 years be identified as predictor for 30-day and for long-term mortality?

2. Patients and methods

This is a retrospective observational file study. The patients were included consecutively and all important preoperative, operative and early postoperative data were found in the electronic files. In the late 1990, fast track anesthesia was introduced and the use of opioid analgesia was reduced. Therefore, some details such as the duration of mechanical ventilation and stay in intensive care unit were recorded only for the last 1500 patients of the entire cohort. Moreover, the electronic accessibility of medical files allowed a good long-term follow-up of the patients with respect to survival. However, long-term follow-up data concerning complications and intercurrent diseases from patients admitted in nursing homes and facilities were mostly restricted to hospital admissions. Especially with development of dementia, data about complications were difficult to retrieve. In some older patient files, not all details could all be retrieved and were left out. This was more the case in patients operated before 2000. This included the Euroscore, because measurement of the pulmonary artery pressure was not a routine before 2000. Furthermore, the NYHA class was not always easy to estimate in elderly patients, while LV ejection fraction could be found in most, but not all patient files. However, the missing details were spread evenly among both age group. With the available data, a meaningful analysis was still possible. We could assume safely that the missing data would not alter significantly the outcome. This population is a part of a larger population of 2500 consecutive patients who underwent SAVR with a biological valve. Of these, 1855 patients were younger than 80 years old. The inclusion criteria were implantation with a biological heart valve (BHV) prosthesis and an age below 80 years. The main exclusion criteria were any BHV prosthesis in another position or a mechanical valve in any position. By including concomitant procedures as listed below, the patient groups showed some heterogeneity, but it also reflected the ‘real world’ situation. The preoperative profile included the factors which were also investigated in prior series [12,14–17]. The preoperative factors under study were chronic kidney dysfunction (CKD), defined by an estimated glomerular filtration rate (GFR) of <60 mL/min, chronic obstructive pulmonary dysfunction (COPD) defined by a forced expiratory volume at 1 second of < 80% of the predicted value, diabetes mellitus and hypertension treated by diet and chronic medication, a history of histopathological documented malignancy, a history of sudden neurologic ischemic event, an atrial fibrillation or a conduction defect documented on ECG, severity of the valve disease and left ventricular function recorded by echocardiography, acute myocardial infarction recorded on ECG and by elevated heart enzymes, endocarditis, defined by the modified Duke criteria, coronary and peripheral artery disease (CAD, resp. PAD) documented as a > 50% lesion on angiography, prior percutaneous or surgical procedures on the coronary arteries, or permanent pacemaker (PPM) implantation, need for urgent SAVR (defined as a need for surgery during the admission at which the diagnosis was made) or emergent SAVR (defined as a need for SAVR within 24 hours). The operative parameters included aortic cross-clamp (ACC) time and cardiopulmonary bypass (CPB) time, expressed in minutes, association of CABG, of mitral valve repair, a procedure on the ascending aorta, or other procedures (maze, closing of a patent foramen ovale). The postoperative factors included the need for resources (the need for blood products, duration of mechanical ventilation, length of stay in intensive care unit and hospital stay) and adverse events, such as delirium with agitation, acute renal injury (a decrease of GFR with at least 25%), clinical signs of atelectasis or pneumonia, confirmed by medical imaging, endocarditis, thromboembolism (sudden neurologic deficit or ischemia of a limb) of any severity, bleeding of any severity, new or recurrent atrial fibrillation, new or progress of a preexistent conduction defect, low cardiac output (need for mechanical support, prolonged intravenous inotropes, low blood pressure with pulmonary or peripheral edema) and mortality. The Euroscore II could be calculated for the last 1000 patients, when pulmonary artery hypertension could be documented by an echocardiographic signal. Statistical methods included a chi-square analysis for categorical variables and a Student’s T-test was performed for continuous parameters. A Cohen’s D (the ratio between the difference of the mean values and the standard deviation) was calculated as a measure for effect size. At a value of 0.8 or larger, the effect size could be considered as large and clinically relevant. At a level of 0.5 this was considered as medium sized, and at a level of 0.200 or lower, this was small and therefore clinically not relevant. For long-term analysis, a Kaplan-Meier analysis with a Log rank test was performed. For the predictors of long-term mortality, a Cox’ proportional hazard analysis was performed. A multivariate logistic regression analysis was performed to identify predictors for 30-day mortality, as alternative for a propensity score match analysis. A multivariate regression analysis is more robust to low exposure prevalence and performs better in large patient sample sizes. Logistic regression is the technique of choice when the numbers of events are high. A p-value of less than 0.05 was considered as statistically significant. This study was approved by the ZNA ethical committee under the protocol number 2656. The database is an joint ownership of the authors.

3. Results

3.1. Association of age class on preoperative and operative factors

Most implanted valves were a Carpentier Edwards (n = 942) and Carpentier-Edwards Magna Ease (n = 699) valves. Mosaic or Mosaic Ultra (n = 80), Mitroflow (n = 52), Crown (n = 47), Perceval (n = 15) and other devices (n = 20) were used in smaller numbers. These valves were evenly distributed amongst both age classes. Table 1 shows the distribution of preoperative and operative factors in both age groups or the mean with the standard deviation. The older age group were more female, had more atrial fibrillation, more CKD, more pulmonary and systemic artery hypertension, more CAD and prior acute myocardial infarction, more conduction defects and a higher symptomatic status according the New York Heat Association (NYHA) classification. The mean ejection fraction of both groups was equally at 61.5%. The mean transvalvular gradient (TVG) was significantly lower in the younger age group. This observation corresponded with a larger aortic valve area. The effect size of both valvular parameters was small (Cohen’s D < 0.200). As could be expected, the mean Euroscore was about 2% higher in the older group, but this effect size was also limited (Cohen’s D of 0.280) and therefore of less clinical relevance. Noteworthy, less than a quarter of the older age group has a higher surgical risk. The older age group underwent significantly more concomitant CABG, but mitral valve repair was more performed in the younger age group. The mean ACC time between the age groups differed only 2.4 minutes. Its effect size was small (Cohen’s D of 0.1038,), and is of limited clinical relevancy. The mean CPB time differed less than a minute.

Table 1.

Distribution of preoperative and operative factors across age groups.

Factor	in younger	in older	p
Preoperative
Male	687/1106 (62.1)	406/749 (54.2)	<0.001
Atrial fibrillation	193/1105 (17.5)	196/747 (26.2)	<0.001
CKD	111/563 (19.7)	137/380 (36.1)	<0.001
Euroscore >8% (high risk)	91/617 (14.7)	98/416 (23.6)	<0.001
Euroscore <4% (low risk)	385/617 (62.4)	210/416 (50.5)	<0.001
Euroscore II as %	4.9 ± 5.7	6.8 ± 8.1	<0.001
Aortic valve area in mm2	77.7 ± 36.1	71.6 ± 24.9	0.002
Pulmonary artery hypertension	145/616 (23.5)	131/427 (30.7)	0.010
Coronary artery disease	641/1106 (58.0)	476/748 (63.6)	0.014
Arterial hypertension	761/1105 (68.9)	553/747 (74.0)	0.016
Mean TVG in mm Hg	43.7+/15.5	46.8 ± 17.4	0.020
Myocardial infarction	151/1103 (13.7)	130/749 (17.4)	0.031
Conduction defects	293/1104 (26.5)	230/745 (30.9)	0.042
NYHA functional class III/IV	564/810 (69.6)	424/568 (74.6)	0.042
Carotid artery disease	196/1106 (17.7)	158/21.2 (21.2)	0.065
Prior PPM implant	28/1106 (2.5)	29/746 (3.9)	0.098
Endocarditis	36/1106 (3.3)	15/748 (2.0)	0.107
Ischemic neurologic event	138/1106 (12.5)	110/748 (14.7)	0.167
LV ejection fraction < 50%	165/837 (19.7)	125/564 (22.2)	0.267
Malignancy	156/1105 (14.1)	117/747 (15.7)	0.358
COPD	273/1083 (25.2)	195/727 (26.7)	0.462
Congestive heart failure	244/1105 (22.1)	176/749 (23.5)	0.475
Diabetes mellitus	247/1106 (22.3)	157/749 (21.0)	0.483
Emergency SAVR	56/1105 (5.1)	34/749 (4.5)	0.603
Peripheral artery disease	184/875 (21.0)	140/635 (22.0)	0.634
Urgent SAVR	146/905 (16.1)	112/677 (16.5)	0.827
Prior CABG	86/1106 (7.8)	60/748 (8.0)	0.847
LVEF as %	61.5 ± 16.4	61.5 ± 15.8	0.958
Operative
Mitral valve repair	52/1106 (4.7)	16/749 (2.1)	0.004
Concomitant CABG	619/1106 (56.0)	461/749 (61.5)	0.017
ACC time in minutes	70.7 ± 24.7	68.3 ± 20.9	0.040
Procedure ascending aorta	100/1106 (9.0)	62/749 (8.3)	0.567
Smallest valve size	22/1106 (2.0)	13/749 (1.7)	0.694
CPB time in minutes	121.7 ± 45.2	121.3 ± 45.2	0.886

CABG: coronary artery bypass graft; CKD: chronic kidney disease; COPD: chronic obstructive pulmonary disease; CPB: cardiopulmonary bypass; LV: left ventricle; NYHA: New York Heart Association; PPM: permanent pacemaker; SAVR: surgical aortic valve replacement; a p-value <0.05 was considered as statistically significant.

3.2. Effect of age class on early postoperative outcome

Table 2 shows the effect of the age group on the need for resources and on postoperative adverse events. There were no significant differences in need for resources between the age groups, except for stay in the ICU, and the need for packed red blood cells. The effect sizes (Cohen’s D) were 0.150, resp. 0.214, which indicated that these differences had little clinical significance. This observation could be related to a borderline lower hematocrit in the older age group (24.7 ± 3.4% vs. 25.2 ± 3.6%, p = 0.052, Cohen’s D of 0.142) and almost no difference in the nadir of thrombocytes below 80,000 (70/611 or 11.5% in the younger group vs. 45/412 or 10.9% in the older group, p = 0.791). The need for reintervention did also not differ between both age groups. Delirium, pulmonary complications, acute renal injury and conduction defects were significantly more present in the older patients. There was only a trend for an increase in PPM implant, and no significant difference in the need for renal replacement therapy or in ventilation time. This indicated that the complications were not severe enough to warrant these resources in a higher degree in the older age group. There was also no significant difference in 30-day mortality. The Euroscore II could only be determined for patients in a later era, but the difference in mortality for those patients also remained non-significant. The Euroscore overestimated the risk with 0.9% in the younger age group and with 1.5% in the older age group. Additionally, the association of CABG had no effect on the immediate outcome in patients younger than 75 years: the 30-day mortality rate for isolated SAVR was 19/619 or 3.1%. In patients with the combined operation, this was 16/487 or 3.3%, which was not significantly different (p = 0.839). In patients older than 75 years, the 30-day mortality rate without associated CABG was 10/289 or 3.5%. In patients with associated CABG, this was higher with 24/460 or 5.2%, but this increase was also not significant (p = 0.261).

Table 2.

Effect of age on need for resources and on early postoperative events.

Factor	in younger	in older	p
Need for resources
Units of packed cells	2.1 ± 3.2	2.9 ± 3.4	<0.001
Postoperative LOS	11.8 ± 7.9	14.4 ± 10.6	0.001
LOS-ICU (days)	2.5 ± 5.5	3.6 ± 9.7	0.008
PPM implant	21/1103 (1.9)	23/749 (3.1)	0.106
Renal replacement therapy	30/1099 (2.7)	29/747 (3.9)	0.167
Plasma derivatives	167/611(27.3)	123/413 (29.8)	0.393
Mechanical ventilation (hours)	14.9 ± 52.3	17.7 ± 52.0	0.409
Thrombocyte concentrate	71/612 (11.6)	46/413 (11.1)	0.819
Reintervention	39/1107 (3.5)	26/750 (3.5)	0.948
Adverse event <30 days
Delirium	57/859 (6.6)	66/635 (10.4)	0.009
Acute renal injury	128/615 (20.8)	115/415 (27.7)	0.010
Postoperative creatinine mg%	1.29 ± 0.99	1.45 ± 1.02	0.017
Increase in mg%	0.28 ± 0.68	0.39 ± 0.76	0.014
Conduction defect	146/1105 (13.2)	129/748 (17.2)	0.017
Bleeding of any severity	49/1105 (4.4)	50/748 (6.7)	0.035
Pulmonary complications	116/1107 (10.5)	102/749 (13.6)	0.039
Atrial fibrillation	411/1105 (37.2)	310/749 (41.4)	0.069
Mortality	35/1106 (3.2)	34/749 (4.5)	0.125
With known ES-II	25/618 (4.0)	22/416 (5.5)	0.346
Thromboembolism	33/1105 (3 .0)	29/748 (3.9)	0.296
Ventricular arrhythmias	37/1105 (3.3)	30/747 (4.0)	0.450
Low cardiac output syndrome	57/1105 (5.2)	44/748 (5.9)	0.501
Endocarditis	2/1105 (0.2)	1/748 (0.1)	0.804

ES-II: Euroscore II; LOS-ICU: length of stay in intensive care unit; PPM: permanent pacemaker; a p-value < 0.05 was considered as statistically significant.

Table 3 shows the independent predictors for 30-day mortality rate. Cardiac and renal factors were the only predictors which could be identified. However, need for emergent SAVR was only identified in the older age group. In this age group, congestive heart failure was the main predictor, followed by a need for emergent SAVR. Chronic kidney disease was the third predictor. Table 4 shows the predictors for the hospital mortality of the entire group. Preoperative congestive heart failure was the most clinically relevant (in terms of odds ratio) and also the most significant. Age over 75 had not been identified as a predictor for hospital mortality for the overall patient group. This factor had an odds ratio of 1.52 and a 95% confidence interval between 1.02 and 2.12, with a p-value of 0.947. The strength of the statistical model in terms of a chi-square value did not improve by adding age over 75 years as potential predictor. These values indicated that age of 75 years had no effect on 30-day mortality.

Table 3.

Preoperative & operative predictors for mortality for both groups separately.

	Older group			Younger group
Factor	OR	95%CI	p	OR	95%CI	p
Congestive heart fail	5.94	2.10-16.8	<0.001	3.80	1.54-9.41	0.004
Emergent SAVR	5.20	1.66-16.3	0.005	–	–	–
CKD	3.42	1.30-8.96	0.013	3.81	1.43-9.48	0.004

CKD: chronic kidney disease; SAVR: surgical aortic valve replacement; a p-value < 0.05 was considered as statistically significant.

Table 4.

Predictors for hospital (30-day) mortality for the entire group.

Predictor	Odds ratio	95%CI	p
Congestive heart failure	3.98	1.94–8.14	<0.001
CKD	3.47	1.75–6.87	<0.001
Incomplete revascularization	3.02	1.33–6.84	0.008
Emergency	3.07	1.28–7.35	0.012
COPD	2.05	1.05–3.98	0.036

CKD:chronic kidney disease; COPD: chronic obstructive pulmonary disease; CPB: cardiopulmonary bypass time; a p-value < 0.05 was considered as statistically significant.

Table 5 shows the long-term survival for both age groups. The five-year survival for both age groups was more than 75%. The mean survival time was 142 (137–147) months for the younger age group and 113 (108–118) months for the older patients, with p < 0.001. This difference was 29 months, which was much less than the difference in mean age at operation (69.7 ± 4.9 years vs. 77.0 ± 1.4 years). This outcome is also illustrated by Figure 1, which shows a favorable outcome for both groups, although lower for patients over 75 years.

Table 5.

Survival for both age groups (p < 0.001).

Survival	younger	n	older	n
5 years	85.6 ± 1.1%	(860)	76.7 ± 1.6%	(517)
10 years	61.8 ± 1.7%	(412)	44.6 ± 2.1%	(213)
15 years	29.3 ± 2.0%	(114)	16.3 ± 2.0%	(41)

n: number at risk; a p-value < 0.05 was considered as statistically significant.

Figure 1.

Effect of age over 75 years on long-term survival after surgical aortic valve replacement.

Note: Blue line: survival of patients younger than 75 years; red line: patients of 75 years and older; the Y-axis represents the cumulative survival with the 1.0 value as 100% (cum: cumulative). The X-axis represents follow-up time expressed in months.

Table 6 shows the predictors for long-term mortality. Age over 75 years, although significant, was not the dominant predictors for this outcome. For both age groups separately, major differences could be detected (Table 7). Most non-cardiovascular factors were predictive only in the younger age group. Pulmonary disease was a common predictor for both groups.

Table 6.

Predictors for long-term mortality.

Predictor	odds ratio	95% CI	p
AV-block	1.62	1.22-2.16	<0.001
COPD	1.54	1.31-1.79	<0.001
CKD	1.49	1.23-1.82	<0.001
Age >75 years	1.48	1.28-1.71	<0.001
Atrial fibrillation	1.35	1.13-1.60	<0.001
Peripheral artery dis.	1.31	1.10-1.56	0.002
Diabetes mellitus	1.31	1.11-1.55	0.002
Malignancy	1.31	1.09-1.59	0.005
Urgent SAVR	1.24	1.01-1.51	0.038

AV: atrioventricular; CI: confidence interval; CKD: chronic kidney disease; COPD: chronic obstructive pulmonary disease; dis.: disease; SAVR: surgical aortic valve replacement; a p-value <0.05 was considered as statistically significant.

Table 7.

Predictors for long-term outcome for both age groups.

	Younger patients			Older patients
Predictor	OR	95% CI	p	OR	95%CI	p
CKD	2.16	1.50-3.10	<0.001	–	–	–
Malignancy	1.84	1.32-2.55	<0.001	–	–	–
Diabetes	1.53	1.15-2.02	0.003	–	–	–
Conc-CABG	1.49	1.12-1.97	0.006	–	–	–
COPD	1.47	1.09-1.98	0.012	1.56	1.27-1.92	<0.001
AV-block	1.84	1.13-2.98	0.014	1.66	1.12-2.45	0.012
Atrial fibrillation	1.43	1.03-1.99	0.033	1.36	1.10-1.70	0.005
Urgent SAVR	–	–	–	1.40	1.08-1.82	0.011

AV:atrioventricular; CI: confidence interval; CKD: chronic kidney disease; COPD: chronic obstructive pulmonary disease; conc- CABG: concomitant CABG; OR: odds ratio; SAVR: surgical aortic valve replacement; a p-value <0.05 was considered as statistically significant.

4. Discussion

Following observations were made with respect to both age classes.

First, patients older than 75 years were more female, and had significantly more cardiovascular comorbid conditions. They had an almost 2% higher Euroscore II and a higher degree of NYHA III or IV. Consequently, patients of the older age group were more in a higher risk class. The valve disease in these patients seemed also more advanced because of a higher mean TVG and a lower aortic valve area. The higher prevalence of coronary artery disease, pulmonary artery hypertension and atrial fibrillation adds to the higher risk score and severity of symptoms. Nevertheless, the left ventricular ejection fraction was identical in both age groups and the need for urgent SAVR was also very similar. Chronic kidney disease was the only non-cardiovascular factor significantly more present in the older age group. Similar observation were also made in other series [18]. Although older patients were more in a higher risk class, i.e., an Euroscore of 8% or more, this remained less than a quarter in patients older than 75 years. The presence of an ejection fraction below 50% in both age classes of the current patient population was at a level of about 20%. This was also observed in the previously investigated octogenarian population. Nevertheless, a history of CHF was currently at a level of 22–23% for both current age groups, while for patients between 80 and 84 years, this was 28% and for patients of 85 years and older, this was almost 50% [12]. In general, a good left ventricular function seem to be common in octogenarian patients who are referred for SAVR [19,20]. The need for an urgent SAVR also increased with age: for both current age groups, this was at a level of 16%, and somewhat higher compared to another series [18]. For patients aged between 80 and 84 years, this was 21% and for patients of 85 years and older, this was 30% [12]. These observations indicate that the physiological cardiac reserve decreases with increasing age, even with a maintained ejection fraction. The importance of the need for an urgent or emergent operation on the postoperative outcome were demonstrated in prior series [19,21], especially with increasing age [12]. The prior and current results indicated that in patients aged under 80 years, age did not affect the capacity of the left ventricle to tolerate the burden of the diseased aortic valve. Once, the age limit of 80 years had passed, and certainly at 85 years, the capacity of the left ventricle to tolerate this burden decreased drastically [12].

Second, concomitant CABG was a commonly associated to SAVR in both age groups, but this procedure was significantly more performed in the older age group. This is related the significantly higher presence of CAD with increasing age. This phenomenon was also observed in other similar series [18], as well as in octogenarian patients [12]. However, the need for a concomitant CABG in our current series was high compared to other series studying elderly patients [20,21]. Other operative parameters such as a procedure on the ascending aorta, an inability to perform a complete revascularization and cardiopulmonary bypass times were comparable for both age groups currently under investigation. The rate of incomplete revascularization rose from 8.1% in patients younger than 75 years to 10% in patients younger than aged between 75 and 79 years. Although this was not significant, a further rise could be observed to 12.9% in previously investigated patients between 80 and 84 years and to 14.8% in patients above 85 years [12]. This indicated that inoperability of coronary arteries increases with age. A diffusely affected coronary artery in elderly could play a major role. Mitral valve repair was significantly more performed in the younger age group, but this procedure was uncommon in series investigating octogenarian patients [12,22,23].

Third, postoperative adverse events such as delirium, acute renal injury, a new or a worsening of a preexistent conduction defect, bleeding and pulmonary complications were significantly more observed in the older group. However, the need for resources such as a permanent pacemaker implant, renal replacement therapy, and prolonged mechanical ventilation did not significantly differ between age groups. The exceptions were the need of packed red blood cells and an increased length of stay. This indicated that the impact of most postoperative adverse events remained limited. However, the parallel analysis of the octogenarian group showed that the complication rates and the need for resources increased gradually with age [12]. In some other series the stroke rate was comparable to that of the current series [20,23]. This event was even lower in patients who underwent isolated SAVR, although an increase of the stroke rate with age was also observed [15,23]. Mortality at 30 days was not significantly different between both age groups and was at a level of 3.2% for the younger age group and at 4.5% for the older age group. For the last 1,500 patients for whom an Euroscore II could be calculated, this was 4.0%, resp. 5.5%. Importantly, age of 75 years was not a predictor of 30-day mortality. The most dominant predictors for this event were prior CHF and CKD. These predictors were also identified in both separate age groups. The need for an emergent SAVR was also identified in the group as a whole and in the older age groups, but not in the younger group. With advancing age, patients tolerated a need for emergent surgery poorly. This poor tolerance was also identified earlier in the octogenarian patients in whom a higher mortality rate was observed [12,22]. This increment in mortality was observed for isolated SAVR: this was 3.5% in the 70–79 year group, 4.5% in the 80–85 year group and 5.7% in patients of 85 years and older. For patients undergoing SAVR with CABG, this was 5.0%, 6.4% and 7.4% [23]. Taken together, the higher mortality rate and lesser tolerance for need of urgent surgery indicated that with aging, physiological cardiac reserves decline, especially after reaching the age of 80 years. Once the age limit of 80 years had passed, the capacity of the left ventricle to tolerate this burden drastically decreased, and valve replacement should not be postponed [12,19,21,22].

Fourth, long-term survival was significantly lower in the older age group, but this difference remained limited. The mean postoperative survival time in the patient group, younger than 75 years, was almost 12 years and its 10-year survival was over 60%, which is high. For patients aged between 75 and 79, the mean survival time was 9.5 years with a 10-year survival approaching 45%. These rates come close the age and gender matched survival of the general population. Furthermore, the preoperative mean difference of age between both age groups (69.7 vs. 77.0 years) was much larger compared to the mean postoperative mean survival time, which differed only 29 months or about 2.5 years. This indicated that SAVR offers the greatest gain of survival in the older age group. Nine predictors for long-term mortality were identified. Age over 75 was the fourth ranking predictor and less important compared to chronic renal or pulmonary dysfunction. All predictors were close in terms of odds ratios and of clinical relevance. Calculating the predictors of long-term mortality for both age groups separately showed a somewhat different picture: both age groups shared only three predictors, namely COPD, atrial fibrillation, and prior AV-block. Need for an urgent SAVR was only in the older group identified as a predictor for long-term survival. This was also observed in several octogenarian series [12,22]. Taken together, this seems to indicate that the age-related decrease of the functional cardiac reserve had a long-lasting postoperative effect.

Few series, however, are available to assess the results of SAVR in patients between 75 and 79 years of age, which can be considered as the “gray zone.” More results for SAVR are available about octogenarian patient series, which showed that an acceptable 30-day mortality and 5-year survival could be achieved [20,22,24]. Nevertheless, TAVI is an effective and safe alternative to SAVR for octogenarian patients since it was associated with lower in-hospital mortality, a lower major in-hospital complication rate, and a lower 30-day readmission rate [25]. This procedure was introduced as alternative treatment of symptomatic aortic valve disease in elderly and high-risk patients. Its use was expanded into mid- and low-risk patients. In a group of patients between 65 and 74 years, short-term mortality, stroke and acute renal injury were low after SAVR and comparable to transfemoral TAVI. Bleeding and delirium were higher after SAVR, while vascular complications and need for PPM implant was higher after TAVI [16]. These results indicated that SAVR remained a viable treatment option. This was especially true for isolated SAVR [26]. Another indication of the viability of SAVR as treatment option is the observation that the mortality rate after isolated SAVR also decreased over time for patients between 75 and 79 years from 5.6% to 3.3% [15]. This temporal trend could be attributed to improved surgical techniques and perioperative care. Although patient age is an important factor in the decision making process for the treatment of degenerative aortic valve disease, frailty, valve durability, comorbid conditions and life expectancy may be a better guide than age alone and should guide for the choice of the best treatment option. Cardiac comorbid conditions which require surgical correction would favor the use of SAVR, while prior cardiac surgery would favor TAVI as treatment option [4,5]. The need for collaborative decision making is demonstrated by the redirection of octogenarian patients from TAVI to SAVR by a heart team, with a superior outcome [13]. The reports concerning the choice of TAVI or SAVR in patients with aortic valve stenosis are conflicting. TAVI performed in patients with a worse preoperative profile and a high surgical risk achieved a similar increase in quality of life and functionality compared to patients undergoing SAVR with a lower preoperative comorbidity [27]. However, this is not a universal observation. Patients who underwent a TAVI procedure had a significantly lower five-year survival and survival freedom from major adverse cardiac events, with a higher rate of permanent pacemaker implantation a more than moderate paravalvular leak, compared after SAVR with as Perceval ® device [28].

There are several reasons to opt for SAVR in this so-called “gray zone” of patient age. The first reason could be the increase of life expectancy in Western societies. Life expectancy in several European countries has increased and is projected to increase further. The expected survival of both males and females aged between 75 and 79 years is over 10 years. The surgical option could deliver a valve for this time span, since the life expectancy of patients over 75 years with severe aortic valve disease seems to be completely restored by SAVR. Survival rates of patients who survive the postoperative period are similar to that of the general population for the same age, sex and region [29].

The second reason was the presence of factors favoring SAVR as treatment option in significant portion of the patients in the current population. These factors included an Euroscore of less than 4%, presence of severe CAD, and the relatively low degree of previous cardiac surgery. In presence of CAD, a PCI after TAVI, especially with self-expandable valves, could be challenging [30]. The choice between PCI before, simultaneous or after TAVI in these patients should be tailored to each individual patient, for which benefits and drawback should be balanced [31]. Performing SAVR with CABG in low risk elderly patients could be an attractive alternative. Moreover, in both age groups, the Euroscore overestimated the currently observed mortality. The Euroscore II overestimated mortality in the current group with 0.9% in the younger age group and 1.3% in older age group. This was also observed in a comparative series between TAVI and SAVR [16], but this observation is not universal. In one series, the Euroscore II predicted the observed mortality after SAVR very well, while for STS, there was a slight underestimation of 30-day mortality. Their population was relatively young and associated procedures as well as urgent or emergent cases were not included [32]. There was also a slight overestimation of 30-day mortality in patients who underwent either isolated SAVR with a Perceval® device or after TAVI [28]. In another series with a mean age close to the current series and predominantly biological prostheses, the EuroSCORE underestimated mortality in patients over 75 years (3.7 ± 4.4% versus the observed mortality of 4.8%). The STS score performed worse for this population, with a calculated value at a level of 2.1 ± 1.5%. However, the observed mortality was 6.6% in the elderly patients, while this was 2.8% for the younger counterparts. The area under the receiver-operator characteristics curve was lower for elderly compared to younger patients. This applied to the EuroSCORE as well as to the STS score, which performed less well in elderly patients. Patients older than 75 years seemed to be at increased postoperative risk, regardless of the calculated risk score [18]. A similar observation was made in another series of patients older than 75 years: the calculated Euroscore II of 3.95 ± 2.93% underestimated the observed mortality of 5.86% [29]. In still another series, the EuroSCORE performed well in patients undergoing SAVR with or without CABG with areas under the curve of 0.761 and 0.819. This was close to the performance of the STS score. The inclusive nature of EuroSCORE II for numerous procedures provided more flexibility than the STS score for complex procedures [33]. The EuroSCORE II should be considered for calculating risk score for complex cardiac surgical patients [33]. On could safely state that increasing scores reflected and increase in observed mortality, something which was also observed in octogenarian patients [20]. The third reason is the development of newer biological valve prostheses. The sutureless (Perceval ®) and rapid deployment (Intuity ® and Intuity Elite ®) valves could shorten the ACC and CPB times considerably. A Perceval device might be useful in patients of 75–79 years who are referred for SAVR [34]. This rapid deployment valve has also been used in our series, but in a mainly older population [14]. The use of these devices also allows a minimal invasive approach such as partial sternotomy [17,28] and right anterior or lateral mini-thoracotomy, which is safe and feasible [28,35,36], but which requires a learning curve. This could improve postoperative outcome in frail and elderly patients, especially if there are anatomical conditions prohibiting the use of TAVI. Hence, a rapid deployment valve could become the main option for patients in the “grey zone” with respect to age and operative risk. This valve could be an alternative for SAVR using conventional sutured valves as well as for TAVI. In one series, procedural outcomes and 30-day mortality were comparable between isolated SAVR with use of Perceval® valves and TAVI. The series also documented the typical advantages and drawbacks for TAVI compared to SAVR. The patients who underwent TAVI had a higher rate of paravalvular leaks, a higher need for permanent pacemaker implantation and more vascular complications as well as a lower 5-year survival and freedom from major adverse cardiac events. In surgical patients, a higher rate of surgical revision, bleeding and need for transfusion were observed. Acute renal injury was comparable between both groups [28]. If the development of a structural valve degeneration after SAVR is anticipated, a redo SAVR might be hazardous since these patients have become older and more frail over time. A valve-in-valve TAVI seemed a good option. The Inspiris Resilia® device is mounted on the same frame as a conventional Carpentier-Edwards pericardial valve. The glycerolized tissue allows dry storage and handling. The valve has an expandable frame for the sizes 19 to 25, which allows placement of a transcatheter valve without fracturing during a uniform, controlled and predictable expansion in case of future degeneration. These devices showed an event-free rate for SVD comparable to those after implantation of the older BHV, including the Carpentier-Edwards Magna Ease. The hemodynamic performance and clinical results at 30 days and at 5 years were comparable [37,38]. This approach allows the surgical option in an earlier phase of life, while other options remain open for future TAVI procedures in patients where this is necessary.

5. Limitations

This is a retrospective observation study, with all its inherent limitations. The possible biases have been mitigated by a consecutive inclusion. Access to detailed digitalized medical files allowed a systematic inclusion of operative and perioperative details, especially for patients operated after 2000. For the patients operated before, data were less detailed, but still sufficient for obtaining meaningful results. The long-term follow-up also occurred through access of medical files. However, when patients were admitted to a nursing or retirement home, medical details were limited, especially in presence of dementia. Changes in surgical techniques were the progressive introduction of warm blood cardioplegia since 2012, being the most important. Partial sternotomy in a number of patients was already adopted in a very early stage. Other major changes in management included the reduction of opioid analgesics and hence reduction of the intubation time from often over one day to 4 to 6 hours. This also shortened the time spent in an intensive care unit. For these reasons, length of stay in ICU and of mechanical ventilation were included only in the last 1500 patients.

6. Conclusions

This report investigated the results of SAVR in patients under 80 years of age, especially of those between ages 75 and 79 years, for whom TAVI could be advocated according the current guidelines. Patients below 80 years seemed to have a cardiac reserve capable to withstand major surgery. Thirty-day mortality rate of patients under 75 years and between 75 and 79 years was low, and age over 75 years was not a predictor for mortality. The need for postoperative resources was mostly comparable for both age groups, indicating that the severity of postoperative complications was similar. Survival of both age groups is favorable, with a five-year survival of over 75%. The gain was greater in the older age group. In areas and countries with restricted reimbursement for TAVI, SAVR is a good option to treat symptomatic aortic valve disease. Earlier studies showed that redirection by a heart team of octogenarian patients from TAVI to SAVR resulted in favorable postoperative outcomes. One can expect that this finding could be replicated in patients between 75 and 79 years. Once the disease has become symptomatic in patients over 75 years, surgery should not be postponed. In summary, patients in the “gray zone age range” between 75 and 79 years still have an adequate resilience to withstand SAVR because of the physiologic reserve they possess. Patients aged between 75 and 79 years have also a relative higher long-term survival benefit from SAVR, and could see a restoration of the life expectancy.

Funding Statement

This paper was not funded.

Article highlights

• Patients with symptomatic aortic valve disease, who are 75 years or older, but younger than 80 years could be labeled as the ‘gray zone’ age range with respect to the choice between surgical aortic valve replacement and transcatheter valve implantation as treatment options.

• Patients in this age range, who were referred for surgical aortic valve replacement, had a higher preoperative comorbidity rate and risk score compared to patients younger than 75 years.

• Patient of the older age group required more associated coronary artery bypass grafting (CABG) but less mitral valve repair, while cardiopulmonary bypass time of both age groups was similar.

• Older patients had a higher rate of postoperative adverse events, but the need for resources for older patients was limited to an increased need for blood transfusion and a longer stay in intensive care unit (ICU), indicating that the severity of postoperative events was largely comparable.

• Thirty-day mortality of both age groups did not differ significantly, and the Euroscore II overestimated the 30-day mortality in both age groups.

• Age over 75 years was not an independent predictor for 30-day mortality.

• Need for emergent surgery was a predictor for 30-day mortality, but only in the older age group, indicating that there might be an age-related decrease of tolerance for the burden imposed by the diseased aortic valve on the left ventricle.

• Long-term survival was favorable for both age groups, but the relative gain was higher in the older group.

• Surgical aortic valve replacement (SAVR) remains a valid option to treat symptomatic aortic valve disease in the age group of 75–79, because it has the capacity to restore age-related life expectancy.

Disclosure statement

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Ethical approval

This study was approved by the ZNA ethical committee under protocol N° 2656.

Individual informed consent was waived by the ethical committee.

Author contributions

ID: data production and collection, analysis, writing and editing

KD: data production and collection, writing and editing

AV: data production and collection, writing and editing

WM: concept and design, data collection and analysis, writing and editing

Data availability statement

The multipurpose database will serve for further publications and will not be shared for the foreseeable future