Effect of Sex and Flow Status on Outcomes After Surgical or Transcatheter Aortic Valve Replacement

Background Low stroke volume index <35 ml/m2 despite preserved ejection fraction (paradoxical low flow [PLF]) is associated with adverse outcomes in patients with aortic stenosis undergoing transcatheter aortic valve replacement (TAVR) or surgical aortic valve replacement (SAVR). However, whether the risk associated with PLF is similar in both sexes is unknown. Objectives The purpose of this study was to analyze the risk associated with PLF in severe aortic stenosis for men and women randomized to TAVR or SAVR. Methods Patients with ejection fraction ≥50% from the PARTNER (Placement of Aortic Transcatheter Valves) 2 and 3 trials were stratified by sex and treatment arm. The impact of PLF on the 2-year occurrence of the composite of death or heart failure hospitalization (primary endpoint) and of all-cause mortality alone (secondary endpoint) was analyzed. Analysis of variance was used to assess baseline differences between groups. Multivariate Cox regression analysis was used to identify predictors of the endpoint. Results Out of 2,242 patients, PLF was present in 390 men and 239 women (30% vs 26%, P = 0.06). PLF was associated with a higher rate of NYHA functional class III to IV dyspnea (60% vs 54%, P < 0.001) and a higher prevalence of atrial fibrillation (39% vs 24%, P < 0.001). PLF was a significant predictor of the primary endpoint among women undergoing SAVR in multivariate analysis (adjusted HR: 2.25 [95% CI: 1.14-4.43], P = 0.02) but was not associated with a worse outcome in any of the other groups (all P > 0.05). Conclusions In women with PLF, TAVR may improve outcomes compared to SAVR. PLF appears to have less impact on outcomes in men.

S ince the inception of transcatheter aortic valve replacement (TAVR) 30 years ago, 1 it has proven to be a noninferior or superior alternative to surgical aortic valve replacement (SAVR) among patients with severe aortic stenosis (AS).
Although short-and medium-term outcomes are comparable between the 2 treatment modalities for patients at low and intermediate risk, 2,3 long-term outcomes for TAVR are still uncertain, and TAVR is therefore primarily used in patients with shorter life expectancy or intermediate/high surgical risk, after individual heart team evaluation. 4,5though AS was previously thought to be similar in men and women, several sexspecific studies have shown that there are differences between the sexes in pathophysiology, epidemiology, presentation of AS, and outcome after aortic valve replacement (AVR). 6Among high-risk patients in the PARTNER (Placement of Aortic Transcatheter Valves) 1 trial, women had significantly reduced 2-year survival with SAVR compared to TAVR. 7Whether this applies to low-and intermediate-risk patients is unknown, but in an analysis of the PARTNER 2, SAPIEN 3 cohort of highand intermediate-risk patients, neither the combined nor individual risk cohorts showed sex differences in mortality or disabling stroke out to 1 year. 8Moreover, patients with AS often present with a low flow state, defined as a stroke volume index (SVi) <35 ml/m 2 , either with reduced (classical low flow) or preserved (paradoxical low flow [PLF]) left ventricular (LV) ejection fraction (LVEF).PLF has been shown to be associated with increased mortality compared to patients with normal flow (NF) after AVR. 9-11PLF is usually more prevalent among women, 10,12 but whether the prognostic implication of PLF is the same in men vs women after AVR is unknown.
The aim of this study was, therefore, to investigate the prognostic implication of PLF among men and women after TAVR and SAVR among patients with low and intermediate surgical risk enrolled in the PARTNER 2 and 3 trials.

METHODS
STUDY SAMPLE.The PARTNER 2 and 3 trials were multicenter, randomized clinical trials comparing TAVR and SAVR among patients with intermediate operative risk (PARTNER 2: Society of Thoracic Surgeons score 4%-8%) 2 and low operative risk (PART-NER 3: Society of Thoracic Surgeons score <4%). 3All patients had severe AS, with aortic valve area (AVA) #0.8 or 1.0 cm 2 in the 2 trials, respectively, and mean gradient $40 mm Hg.Patients in PART-NER 2 were randomly assigned to SAVR with any commercially available bioprosthetic surgical valve or TAVR with an Edwards SAPIEN XT, in PARTNER 3 to SAVR with any commercially available bioprosthetic surgical valve or TAVR with an Edwards SAPIEN 3 transcatheter heart valve.Full lists of inclusion and exclusion criteria have been previously published. 2,3The PARTNER trials were approved by the Institutional Review Board at each site, and written informed consent was obtained from all the patients.
In the present analysis, we excluded patients with reduced LVEF <50%, as well as patients with missing data for AVA, SVi, mean gradient, or LVEF.
ECHOCARDIOGRAPHY.Echocardiograms were analyzed at a core laboratory blinded to randomization and clinical endpoints, as previously described. 13LV dimensions and mass, as well as biplane Simpson LVEF, were estimated as recommended by the American Society of Echocardiography guidelines for chamber quantification. 14Peak aortic jet velocity was measured at the window of maximal velocity, and mean gradient was estimated using the simplified Bernoulli equation.Stroke volume was calculated by multiplying the LV outflow tract diameter and Doppler velocity time integral measured just proximal to the native annulus.AVA was calculated from the continuity equation and indexed to body surface area (AVAi). 15Valvular regurgitations at baseline were assessed and graded as recommended by the American Society of Echocardiography. 16Patients were dichotomized according to guidelines as NF: SVi $35 ml/m 2 or PLF: SVi <35 ml/m 2 . 5tient-prosthesis mismatch (PPM) was defined using the normal reference value of AVA (for the model and size of the implanted prosthesis) indexed to the patient's body surface area, with thresholds adjusted for obese patients as per guidelines. 17stoperative echocardiograms performed 1 month after AVR were analyzed for the presence of transvalvular or paravalvular regurgitation and graded according to American Society of Echocardiography guidelines. 16Mean gradient, maximal jet velocity, and AVA were also measured at 1 month.
PROCEDURE.Details on the procedure for implantation of the Edwards SAPIEN XT 18 and the SAPIEN 3 19 TAVR systems have been previously published.Procedures were performed via transfemoral or alternative access, depending on preprocedural assessment.
Dual antiplatelet therapy with clopidogrel and aspirin was recommended for at least 1 month after TAVR.Carter-Storch et al

Sex and Flow Impact After AS Intervention
M A R C H 2 0 2 4 : 1 0 0 8 5 3 STUDY ENDPOINTS.The primary endpoint in this substudy was the composite of death from any cause or rehospitalization for heart failure symptoms within 2 years.The secondary endpoint was all-cause mortality within 2 years.Time to event was defined as the time to the first endpoint within 2 years; for censored cases, time to event was either the time to the last participation date, or 730 days, whichever was shorter.A clinical events committee independently adjudicated all potential events.Clinical outcomes were reported as defined by the Valve Academic  2).For TAVR patients, similar findings were observed with 9.6% of women and <0.1% of men receiving a 20 mm valve size, and again women with PLF more often received smaller valves compared to women with NF (Table 3).
Women had longer hospital stays after the procedure and were less frequently discharged to their homes.The 30-day mortality rate for the whole sample was numerically higher without statistical significance among women with NF and PLF (2.2 and 2.5%) compared to men with NF and PLF (1.2 and 1.3%) (ANOVA P ¼ 0.26).
At the 1-month echocardiogram, patients with PLF had smaller AVAi despite lower mean gradients, and more often had patient-prosthesis mismatch (Tables 2   and 3).This was most pronounced for women with PLF undergoing SAVR, where 63 (62%) had patientprosthesis mismatch (Table 2).

ASSOCIATION OF PLF AND SEX WITH OUTCOMES.
During a median follow-up of 2 (IQR: 2-2) years, there were 180 deaths and 103 hospitalizations for heart failure.
In the whole sample, significant predictors of the primary endpoint were age, atrial fibrillation, chronic obstructive pulmonary disease, lower LVEF, and lower mean gradient prior to the procedure.In men, the same variables, except LVEF, were predictors of the endpoint, while in women atrial fibrillation was the only significant independent predictor (Figure 1).
Low preprocedural SVi was not a risk factor for 2-year all-cause mortality in the whole cohort, or in

DISCUSSION
This analysis of data from the PARTNER 2 and 3 trials represents one of the largest studies on the impact of sex and preprocedural flow status on outcomes following SAVR and TAVR.Our main findings are that in patients with severe AS and preserved LVEF: 1) PLF was a risk factor for death or heart failure hospitalization among women undergoing SAVR but not TAVR; 2) women with PLF had a trend toward increased risk of all-cause mortality with SAVR compared with TAVR; and 3) PLF was not associated with worse outcomes among men undergoing either SAVR or TAVR.Reduced SVi is a well-described risk factor for adverse outcomes in AS in observational studies, 12,[21][22][23][24][25] although this association is attenuated in multivariable adjustment due to differences in comorbidities. 12,25The increased risk of events is, in part, related to adverse LV remodeling with higher relative wall thickness, 9,10,24 more advanced diastolic dysfunction, 24,26 and a higher proportion of patients with atrial fibrillation.whereas in our study, we show that the impact of PLF is not uniform among men and women: that is, TAVR appears to be superior to SAVR in the vulnerable subset of women with PLF, whereas women with NF or men (regardless of flow status) have similar outcomes with SAVR and TAVR.This finding is consistent with a previous study showing that PLF is an independent risk factor for long-term mortality after SAVR, but that this risk factor was only present in women or more prevalent in women. 9It highlights the need to stratify AS studies according to sex, because of sex-specific differences in pathophysiology and prognosis.
The PLF group in this study had a higher proportion of diabetes, atrial fibrillation, and worse baseline symptom severity compared to the NF group, all of which are risk factors for worse prognosis after AVR; 10,13 these factors may explain, at least in part, the association between PLF and outcomes.However, after multivariable adjustment for known risk factors, PLF remained an independent risk factor among women undergoing SAVR.
The second interesting finding in this article was that among women with PLF, SAVR was associated with a higher 2-year mortality compared to TAVR.This should be interpreted with caution as the number of events in this subgroup was small, and because the association was no longer significant after multivariable adjustment.According to current guidelines from the American College of Cardiology/American Heart Association, both SAVR and transfemoral TAVR can be considered equal treatments and the choice depends on expected longevity of the patient and valve durability. 4The guidelines do not take into consideration sex and flow status in the decision process for the management of patients with AS.However, based on our study, PLF could tip the decision in favor of TAVR compared with SAVR for women.
13,27 In PLF patients in the PARTNER 1 trial, SAVR resulted in an early increased hazard compared to TAVR, though longterm survival was similar. 11This early increase in mortality and in-hospital complications among PLF patients was confirmed in another observational study of patients undergoing SAVR, where known risk scores underestimated the perioperative risk in this group. 10Part of the increased surgical risk associated with PLF may be attributable to concomitant risk factors, explaining why the association between PLF and increased long-term mortality after SAVR was no longer significant after multivariable adjustment. 109][30] Furthermore, women have higher 30-day mortality and higher in-hospital complication rates after SAVR than men, even after propensity score matching. 29,31Several factors may contribute to this increased risk in women vs men.In the present study, the prevalence of frailty was higher in women than in men.Frailty has been reported as an independent predictor of outcome after AVR. 32,33[36] Women also have smaller aortic annuli than men, 12,37,38 increasing their risk for prosthesis-patient mismatch, which may hinder the reverse LV remodeling process, regression of LV diastolic dysfunction, and thus prognosis following AVR. 39,40This issue may be more important with SAVR than TAVR, because of the higher risk of severe prosthesis-patient mismatch associated with SAVR in patients with a small annulus. 40,41In the SAVR arm of the present study, women, especially women with PLF, received smaller valves than men, and they had the lowest AVA, indexed AVA, and the highest proportion of PPM at 1 month.Better effective orifice areas, lower postoperative gradients, and lower incidence of PPM achieved with TAVR may explain, at least in part, the better prognosis associated with TAVR vs SAVR in women.
STRENGTHS AND LIMITATIONS.By combining data from the PARTNER 2 and 3 studies, we were able to perform one of the largest studies on preprocedural PLF in patients randomized to SAVR or TAVR.Previous studies have mostly been performed on retrospective data with the risk of residual confounding despite multivariable adjustment.Nevertheless, there are some limitations to this study.The first is that enrolled patients had to fulfill the inclusion criteria for the randomized trials, and thus our results may not be generalizable to all patients with AS.
We excluded patients with LVEF <50% and our conclusions can therefore not be extrapolated to patients with classical low flow.Furthermore, in the PARTNER 2 and 3 trials, patients with mean gradient <40 mm Hg measured at rest at the recruitment sites were excluded, and our cohort therefore mostly consists of patients with low flow, high gradient AS.Although a number of patients with mean gradient <40 mm Hg measured by the core lab are present, the sickest PLF patients (ie, the patients with the lowest gradients) were not included in the PARTNER trials.However, as a significant proportion of patients with PLF, low gradient according to core lab measurements are present in this study and given that the difference in outcomes should be more pronounced in sicker patients, we believe our results are applicable to all women with low flow and either high or low gradient.In men, the absence of difference could be debatable and different in patients with more pronounced low gradient.Also, we did not test the possibility of sex-specific SVi thresholds 9 to define low flow, due to the number of studied groups.
We were only able to include 2-year outcomes, so long-term outcomes are still unknown, and should be studied.Furthermore, patient-centered outcomes and functional outcomes were not analyzed, and could be a topic for further studies.
Finally, as mentioned previously, the number of deaths was low in the sample, especially when looking at smaller subgroups, and our results regarding higher mortality for SAVR vs TAVR for women with PLF should therefore be considered hypothesis generating.

CONCLUSIONS
In the PARTNER 2 and 3 trials, PLF pattern was associated with worse outcomes in women undergoing SAVR but not TAVR.Moreover, among women with PLF, SAVR may possibly be associated with higher rates of 2-year mortality compared to TAVR.In men undergoing either SAVR or TAVR, PLF was not associated with worse outcomes.
These findings suggest that even among patients with preserved LV function, sex and flow status should be taken into account in the decision-making process between TAVR vs SAVR and that TAVR may potentially be preferred over SAVR in women with PLF.Further randomized studies are needed to confirm the superiority of TAVR over SAVR in this particular subset of patients.

A
B B R E V I A T I O N S A N D A C R O N Y M S AS = aortic stenosis AVA = aortic valve area AVAi = aortic valve area index LV = left ventricle/ventricular LVEF = left ventricular ejection fraction NF = normal flow PLF = paradoxical low flow SAVR = surgical aortic valve replacement SVi = stroke volume index TAVR = transcatheter aortic valve replacement

FIGURE 1
FIGURE 1 Forest Plot of Significant Multivariable Predictors of Death or Rehospitalization for Heart Failure in Whole Sample

FIGURE 2
FIGURE 2 Cumulative Incidence of Mortality

TABLE 1
Clinical and Echocardiographic Characteristics According to Sex and Flow Profile Values are mean AE SD, median (Q1, Q3), or n (%).ANOVA was used for continuous variables.Chi-square tests were used for categorical variables.An expanded table is presented as in Supplemental Table3.AV ¼ aortic valve; KCCQ ¼ Kansas City Cardiomyopathy Questionnaire; LV ¼ left ventricular; NYHA ¼ New York Heart Association Class dyspnea symptoms; STS ¼ Society of Thoracic Surgeons; SVi ¼ stroke volume index.

TABLE 2
Procedural Information by Sex and SVi in SAVR Group Values are n (%), median (Q1, Q3), or mean AE SD.ANOVA was used for continuous variables.Chi-square tests were used for categorical variables.AV ¼ aortic valve; ICU ¼ intensive care unit; PPM ¼ patient-prosthesis mismatch; SAVR ¼ surgical aortic valve replacement; SVi ¼ stroke volume index.

TABLE 3
Procedural Information by Sex and SVi in TAVR Group Values are n (%), median (Q1, Q3), or mean AE SD.ANOVA was used for continuous variables.Chi-square tests were used for categorical variables.AV ¼ aortic valve; ICU ¼ intensive care unit; PPM ¼ patient-prosthesis mismatch; SVi ¼ stroke volume index; TAVR ¼ transcatheter aortic valve replacement.
Sex and Flow Impact After AS Intervention Abbott.Dr Pibarot has received funding from Edwards Lifesciences, Medtronic, Pi-Cardia, and Cardiac Phoenix for echocardiography core laboratory analyses and research studies in the field of transcatheter valve therapies, for which he received no personal compensation; and has received lecture fees from Edwards Lifesciences and Medtronic.Lang RM, Badano LP, Mor-Avi V, et al.Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging.J Am Soc Echocardiogr.2015;28:1-39.e14.Zoghbi WA, Asch FM, Bruce C, et al.Guidelines for the evaluation of valvular regurgitation after percutaneous valve repair or replacement: a report from the American Society of Echocardiography developed in collaboration with the Society for Cardiovascular Angiography and Interventions, Japanese Society of Echocardiography, and Society for Cardiovascular Magnetic Resonance.J Am Soc Echocardiogr.2019;32:431-475.17. Hahn RT, Leipsic J, Douglas PS, et al.Comprehensive echocardiographic assessment of normal transcatheter valve function.J Am Coll Cardiol Img.2019;12:25-34.18. Webb JG, Altwegg L, Masson JB, Al Bugami S, Al Ali A, Boone RA.A new transcatheter aortic valve and percutaneous valve delivery system.J Am Coll Cardiol.2009;53:1855-1858.19.Kodali S, Thourani VH, White J, et al.Early clinical and echocardiographic outcomes after SAPIEN 3 transcatheter aortic valve replacement in inoperable, high-risk and intermediate-risk patients with aortic stenosis.Eur Heart J. 2016;37: 2252-2262.20.Kappetein AP, Head SJ, Généreux P, et al.Updated standardized endpoint definitions for transcatheter aortic valve implantation: the Valve Academic Research Consortium-2 consensus document.J Am Coll Cardiol.2012;60:1438-1454.21.Dayan V, Vignolo G, Magne J, Clavel MA, Mohty D, Pibarot P. Outcome and impact of aortic valve replacement in patients with preserved LVEF and low-gradient aortic stenosis.J Am Coll Cardiol.This study suggests that TAVR should be preferred over SAVR in women with PLF.Further randomized studies are needed, however, to confirm this finding.