Outcomes of emergency compared to elective TAVI: a meta-analysis

INTRODUCTION

Aortic stenosis (AS) is the most commonly treated valvular heart disease in Western countries.1 In a relatively small but growing proportion of patients (from 3.5% to 12%), AS may present as cardiogenic shock (CS) with an estimated short-term mortality as high as 70% if definitive surgical or percutaneous treatment is not provided.2 CS is characterized by an inadequate tissue perfusion as a result of a decompensated cardiac disease that translates into a low-output state. Early management is directed toward keeping a steady hemodynamic profile and ensuring tissue oxygenation through medication or advanced support.3 However, specific therapies are required to ensure a complete resolution, yet conventional surgical aortic valve replacement (SAVR) is often associated with a very high risk of mortality.2

Several trials have shown that transcatheter valve implantation (TAVI) is a safe alternative to SAVR in low-to-high risk patients in stable situations and it is currently considered the preferred alternative in those of high prohibitive surgical risk.4-7 Nevertheless, the risk scores for the main studies that settled the evidence for TAVI procedures were estimated after excluding patients with CS. As a consequence, the main outcomes in this challenging scenario have not been randomly compared to surgery. Actually, such a comparison is unlikely to be performed due to the highly variable baseline profile and differential availability of resources such as mechanic circulatory assist devices. In addition, the different outcomes in emergency TAVIs and planned interventions have been scarcely researched; still, they are key to improve results in what stands as the worst possible clinical scenario. We aimed to assess the current outcomes of emergency/urgent TAVI and the main factors conditioning its prognosis through a systematic review and meta-analysis.

METHODS

Literature search strategy

A systematic review of all published articles in PubMed and Google Scholar databases between January 2014 and January 2020 regarding emergency/urgent versus elective TAVI in severe AS was independently performed by 2 of the authors (A. Aparisi and M. Carrasco-Moraleja). Searched terms were “emergency” and/or “urgent”, “elective”, AND “transcatheter valve replacement” or “TAVR” (transcatheter aortic valve replacement) or “heart failure” and/or “cardiogenic shock”. Definition of emergency/urgent procedures was variable, but the consensus reached for this article was to include patients who required an unscheduled TAVI procedure to treat their refractory heart failure or CS to correct this condition within the next 72 hours after admission. A total of 7 studies8-14 were chosen, and the inclusion criteria established by our group were: a) the study population included patients with aortic stenosis who underwent TAVI; b) only cohort studies that compared emergency or urgent to elective TAVI were included; c) only full English peer-reviewed papers with enough data of outcomes were chosen. The selected exclusion criteria were: a) abstracts; b) case reports; c) editorials; d) experts’ opinions; and e) repetitive studies. Discrepancies between reviewers were resolved through discussion, and consensus was reached. Flowchart is shown on figure 1 and the main features of the studies included are shown on table 1 of the supplementary data.

Figure 1. Flowchart showing search results and selection of the studies included in the meta-analysis.

Table 1. Baseline clinical and echocardiographic characteristics of patients undergoing elective or emergency TAVIs

Primary endpoints

Primary endpoints were short-term mortality and procedural success. Secondary outcomes were perioperative complications. Complications were mostly reported by using the definitions established by the Valve Academic Research Consortium-2.15

Statistical analysis

Qualitative variables were expressed as absolute frequency and percentage. Continuous variables were expressed as mean ± standard deviation unless specified otherwise. To compare the demographic variables and the risk factors between thew groups, the chi-square test or Fisher’s exact test were used for the categorical variables. The Student t test was used for the continuous variables, when applicable.

As a measure of the combined effect, the studies included the odds ratio (OR), a 95% confidence interval, and statistical significance. The homogeneity among the studies was compared using the QH statistic. With regard to the low sensitivity of this test, P values < .10 were considered significant. To somehow overcome this limitation, the I2 statistic was estimated as well, which measures the percentage of the overall variation of the studies explained by the heterogeneity and its 95%CI. A random effects model was used for cases in which the I2 statistic was > 50% and a fixed effects model was used for the opposite cases. The potential publication bias was assessed using a funnel plot, Egger’s test, and Begg and Mazumdar rank correlation test. In the presence of publication bias, the trim-and-fill method was used to reassess the pooled OR. Sensitivity analyses sequentially eliminating dissimilar studies were also conducted.

All P values were 2-tailed. Statistical analyses were conducted using the R software, version 3.6.1 (R Project for Statistical Computing) and Review Manager 5.3.

RESULTS

Patient distribution and baseline characteristics

Seven studies were selected including a total of 84 427 patients who underwent TAVIs, with 70 186 elective procedures (83.1%) and 14 241 emergency ones (16.9%). The main baseline characteristics according to the elective or emergency character of the intervention are shown on table 1 and sensitivity and asymmetry analyses are shown on table 2 of the supplementary data and figure 1 of the supplementary data. Asymmetry was detected for acute kidney injury and, therefore, the trim-and-fill method had to be used to reassess the odds ratio. The percentage of men who underwent elective procedures (52.1%) was higher compared to emergency interventions (50.27%, P < .001). Overall, patients treated urgently showed more comorbidities as summarized by the logistic EuroSCORE (65.9% ± 21% vs 29.4% ± 18%, P < .001) and the Society of Thoracic Surgeons Risk Score (STS) (29.4 ± 27.4 vs 13.7 ± 11.6, P < .001). However, the classical cardiovascular risk factors did not differ among groups (hypertension and diabetes mellitus) and the rates of myocardial infarction and percutaneous coronary intervention were similar. On the contrary, those treated urgently more often had undergone a previous aortic valve replacement. Regarding the main echocardiographic characteristics, emergency procedures were performed in patients with left ventricular (LV) function deterioration (39.5% ± 17.8% vs 52.5% ± 12.8%; P < .001), larger LV end-diastolic diameters (55 ± 9 vs 48 ± 7; P < .001), but similar aortic valve areas (0.66 ± 0.21 vs 0.70 ± 0.23; P = .308), and transaortic mean gradients (40.3 ± 18.3 vs 43.9 ± 16.3; P = .061).

Table 2. Procedural characteristics of patients undergoing elective or emergency/urgent TAVIs

Perioperative characteristics

Procedural results from the studies included are shown on table 2. Transfemoral access (79.3% vs 76.8%; P = .177) and use of general anesthesia (84.5% vs 85.4%; P = .17) were the preferred approaches in both groups. Elective TAVIs showed a higher procedural success rate (93.6% vs 92.5%; P = .007) and a lower need for mechanical circulatory support (1.96% vs 3.56 %; P < .001). Other procedural outcomes were comparable between both cohorts.

Postoperative outcomes

The main postoperative outcomes are shown on table 3 and figure 2. The ORs for perioperative myocardial infarction, life-threatening bleedings, need for permanent pacemaker implantation, and stroke were similar regardless of the planned or emergency setting. On the contrary, the elective cohort showed a smaller rate of acute kidney injury (9.6% vs 22.4%; OR = 2.26; 95%CI, 1.84-2.76; P < .001), and need for dialysis (1.1% vs 2.8%; OR = 2.37; 95%CI, 2.09-2.68; P < .001). Overall, this translated into shorter hospital stays for elective cases, lower in-hospital (3.3% vs 5.75%; OR = 1.32; 95%CI, 1.32-2.83; P < .001), 30-day (4.43% vs 8.84%; OR = 3.13; 95%CI, 1.68-5.80; P < .001), and 1-year mortality rates (19.7% vs 34.47%; OR = 2.87; 95%CI, 1.67-4.94; P = .0001) for elective TAVI (figure 3).

Table 3. Main postoperative outcomes of patients undergoing elective or emergency/urgent TAVIs

Figure 2. Forest plot showing the main postoperative complications of patients included in the meta-analysis.* * Vertical line represents “no difference” point between the emergency and the elective TAVI groups. Horizontal lines represent the 95%CI. Squares represent the OR for each study (the size of each square shows the amount of information given by each study). Diamonds represent pooled OR from all studies.

Figure 3. Forest plot showing the in-hospital to 1-year mortality rates of patients included in the meta-analysis.* * Vertical line represents “no difference” point between the emergency and the elective TAVI groups. Horizontal lines represent the 95%CI. Squares represent OR for each study (the size of each square shows the amount of information given by each study). Diamonds represent pooled OR from all studies.

DISCUSSION

When patients with AS present with severe acute heart failure (HF) or CS the 5-year all-cause mortality is above 60% despite the implementation of therapies to treat valvular heart disease, which poorly compares to this rate in patients free of HF (~20%) or with chronic HF symptoms (~30%) in this setting (16). Determining the factors that condition such a high mortality rate is the key to improve the management of this growing group of patients. The main findings of this study are: a) patients who required emergency TAVIs had a higher baseline risk compared to planned procedures, not only due to the emergency setting, but also to a high burden of comorbidities and deterioration of LV function; b) although procedural success rate was significantly higher in planned cases, this difference was small (93.6% vs 92.5%; P = .007) suggestive that the higher short- and mid-term mortality rates seen in emergency cases were mainly due to postoperative complications, not to the intervention per se; c) need for mechanical circulatory support and dialysis was higher after emergency cases. The early identification of patients at risk who may require these therapies might be useful for a better indication of TAVI in emergency settings.

Baseline characteristics and predicted mortality

In our study, emergency/urgent TAVI patients had a more significant number of comorbid conditions compared to those who underwent elective procedures. We should mention that the Society of Thoracic Surgeons (STS) score has been widely used to assess mortality risk in SAVR patients.17 Nevertheless, the score developed by the Transcatheter Valve Therapy (TVT) group to evaluate in-hospital and 30-day mortality rates18 may be more accurate. According to that score, the prognosis is strongly influenced by the presence of chronic kidney disease (CKD), chronic obstructive pulmonary disease (COPD), and need for emergency TAVI. Of note, AS with concomitant CKD has been linked to higher all-cause and cardiovascular mortality rates compared to patients with AS and without this condition; indeed, the increase of all-cause mortality exponentially correlates with a decline in the glomerular filtration rate.19 In addition, the higher rate of anemia20 and increased bleeding risk in CKD patients is well-known, which conditions a higher need for red blood cell transfusion21 parallel to a deterioration of renal function and survival rate.

The LV function is a well-known prognostic factor of valvular heart disease and its deterioration conditions surgical or transcatheter aortic valve treatment even in asymptomatic patients.22 We should mention that the similar transaortic gradient, despite a reduced LV function in the emerging baseline cases, suggests a more severe valve disease, probable more calcified and degenerated native valves, as also suggested by the higher rate of aortic regurgitation of this cohort. Therefore, the multidisciplinary and multi-imaging approach might be particularly useful for procedural planning and outcome improvement.23

Procedural complications and mortality

Most procedural complications were similar in elective and emergency TAVIs. Although this may be partially explained by the growing operators’ experience worldwide and the lack of differences in the rate of transfemoral approach,24 the greater use of mechanical circulatory support devices may have been particularly relevant in emergency/urgent cohorts. Indeed, the more limited LV contractile reserve of this group of patients can lead to rapid deterioration in the presence of complications like periannular shunts, severe aortic regurgitation or coronary obstruction. Therefore, the presence of risk factors for these complications may suggest the need for circulatory support devices in certain cases before valve implantation as a potential strategy to avoid dreadful prognoses if they occur in the emergency setting.25-27 Prior experience with the Impella device and extracorporeal membrane oxygenation is shown on table 3 of the supplementary data; however, whether there are mortality differences between those with and without mechanical support requires further research. Since procedural success was similar to that of the standard setting, the clinical translation of this is that, even if these cases can be performed successfully in all centers by implanting TAVI, this profile of patients should only be treated in centers with mechanical circulatory support devices available (particularly ECMO), which would exclude low volume or non-surgical centers.

In the present meta-analysis, the cases treated with isolated balloon aortic valvuloplasty were not included. This strategy bears a class IIb-C level of evidence in the last iteration of the guidelines, but it is often used as a bridging therapy to definitive TAVI in hemodynamically unstable patients.28,29 A single-center retrospective study found that TAVI may be superior to a stand-alone balloon aortic valvuloplasty and medical therapy in patients with severe AS and CS, since the isolated balloon aortic valvuloplasty is not free of complications (~25%) and has higher mortality rates.30 Despite of this, large randomized controlled trials exploring this scenario with TAVI are lacking.

Postoperative complications associated with a higher mortality rate

In this systematic review and meta-analysis, we found that emergency/urgent TAVIs had a significantly higher rate of AKI, hemodialysis, and mortality. This is consistent with previous reports that found that patients with post-TAVI AKI were more likely to die. Besides, AKI is a predictor of sepsis, which is also an independent predictor of mortality. The main factors increasing the risk of AKI include CKD, peripheral artery disease, diabetes mellitus, and deterioration of LV function.31,32 A prophylactic strategy may vary from simple hydration with a normal saline solution to forced diuresis with early supportive measures;33 indeed, the use of prophylactic dialysis has been explored in TAVI patients with a high risk of AKI and may be particularly useful in the emergency setting.

Study limitations

There are several limitations related to this systematic review and meta-analysis. First, the studies included were observational since no multicenter randomized studies specifically addressing this topic could be found. Secondly, the definition of emergency/urgent procedures was variable in the studies although an inclusive definition was reached by the study team. Finally, the results may not be generalizable and should be interpreted with caution due to the high heterogenicity reported, which may relate to variability in the study samples and designs.

CONCLUSIONS

In conclusion, the association between emergency/urgent TAVIs and a higher short-to-mid-term mortality rate is mainly due to a high-risk baseline profile, advanced stage of the cardiac disease, and higher rate of acute renal failure. The early identification and referral of patients at high risk for circulatory collapse or AKI need to be properly identified to reduce the TAVI related mortality rate. Further research is needed to elucidate the role of TAVI in emergency or urgent scenarios.

FUNDING

No funding to declare.

CONFLICTS OF INTEREST

I. J. Amat-Santos is a proctor for Boston Scientific.

MATERIAL ADICIONAL

REFERENCES