Iatrogenic atrial septal defects after transseptal puncture for percutaneous left atrial appendage occlusion and their hemodynamic effects

Ioannis Drosos ?,Roberta De Rosa ,Philipp C.Seppelt ,Sebastian Cremer ,Silvia Mas-Peiro2,3 ,Katrin Hemmann ,Jana OppermannRecha Blessing,Mariuca Vasa-Nicotera,Andreas M.ZeiherZisis Dimitriadis

1.Division of Cardiology,Department of Medicine III,University Hospital Frankfurt,Goethe University Frankfurt am Main,Frankfurt,Germany;2.German Center for Cardiovascular Research (DZHK),Berlin,Germany;3.Cardiopulmonary Institute (CPI),Frankfurt,Germany;4.Center of Cardiology,University Medical Center Mainz,Mainz,Germany;5.Division of Cardiology and Intensive Care,Department of Internal Medicine,St.Ansgar Hospital,H?xter,Germany

ABSTRACT Background Percutaneous left atrial appendage occlusion (LAAO) requires puncture of the interatrial septum.The immediate hemodynamic effects of iatrogenic atrial septal defects (iASD) after LAAO have not been examined so far.We aimed at evaluating these effects through invasive measurements of pressure and oxygen saturation.Moreover,we assessed the incidence of persistent iASD at three months.METHODS Forty-eight patients scheduled for percutaneous LAAO were prospectively included in the study.Pressure and oxygen saturation were measured (1) in the right atrium (RA) before transseptal puncture,(2) in the left atrium (LA) through the transseptal sheath after transseptal puncture,(3) in the LA after removal of introducer sheath,and (4) in the RA after removal of introducer sheath.Transesophageal echocardiography was performed at three months to detect iASD.RESULTS Pressure in the RA increased significantly after removing the introducer sheath (P=0.034),whereas no difference was found in oxygen saturation in the RA (P=0.623).Pressure measurement in the LA showed no significant difference after removing the introducer sheath (P=0.718).Oxygen saturation in the LA also showed no significant difference (P=0.129).Followup transesophageal echocardiogram at 3 months revealed a persistent iASD in 4 patients (8.5 %).CONCLUSIONS Our study suggests that iASD after percutaneous LAAO does not result in significant shunts directly after the procedure,although a significant increase of mean right atrial pressure can be observed.Persistent iASDs after percutaneous LAAO seem to be relatively rare at three months.

Reducing risk of stroke represents a milestone in the management of patients with atrial fibrillation.The importance of stroke risk prevention is increasing worldwide,as population age and prevalence of atrial fibrillation are increasing.[1]According to current guidelines,oral anticoagulation (OAC),preferably with non-vitamin K antagonist oral anticoagulants (NOACs),should be considered in patients with a CHA2DS2-VASc score of 1 (or 2 for females) and is recommended in patients with a score of ≥ 2 (or ≥ 3 for females).[2]Guidelines also recommend an early,structured,riskscore-based assessment of bleeding risk,when initiating OAC.[2]Percutaneous left atrial appendage occlusion (LAAO) represents an alternative strategy for stroke prevention,which should be considered in patients with increased bleeding risk or absolute contraindications to OAC.[2,3]Randomized clinical trials and meta-analyses have shown that percutaneous LAAO represents a safe and efficient alternative solution for such cases,which is noninferior to vitamin K antagonists and NOACs.[4-7]Moreover,some recent studies comparing percutaneous LAAO with OAC using propensity score matching have shown a clinical benefit over medical therapy at long-term follow-up,particularly as far as safety and mortality are concerned.[8,9]

Implantation of LAAO devices requires a femoral venous access and puncture of the interatrial septum,in order to enable advancement of the occlusion device to the left atrium appendage.After puncture of the fossa ovalis using a transseptal sheath and needle and confirmation of left atrial position by left atrium pressure tracing,an introducer sheath must be advanced through the septum into the left atrium.Introducer sheaths used for LAAO are mostly 12-14 Fr as opposed to the thicker guiding catheters required for percutaneous mitral valve reconstruction,such as MitraClip procedures,which are as large as 24 Fr.Persistent iatrogenic atrial septal defects (iASD) after MitraClip procedure have been reported in 50% of patients after 6 months[10]and in 27% of patients after 12 months.[11]In the case of ablation therapy for atrial fibrillation,reported rates of persistent iASDs vary,depending on ablation technique and thickness of electrophysiological catheter.[12,13]Literature about the incidence of persistent iASDs in patients undergoing percutaneous LAAO is sparse.Incidence rates of 37% at one month and 11% at 6 months have been reported.[14,15]To the best of our knowledge,there are no studies examining the immediate hemodynamic effects of iASD after percutaneous LAAO.

Our study aims at evaluating possible hemodynamic effects of an iASD after completing the procedure and removing the transseptal sheath.Moreover,we assessed the incidence of iASD directly after percutaneous LAAO and at three months.

METHODS

Study Patients

Patients scheduled for percutaneous LAAO between February and July 2021 were prospectively included in the study.Patients with an atrial septal defect prior to LAAO (either iatrogenic or congenital) were excluded from the study.Our study complied with the declaration of Helsinki and its later amendments and was approved by the Ethics Committee of the University Hospital of Frankfurt (approval number #20-993).Informed consent was obtained from all patients prior to their inclusion in the study.

Echocardiography

All patients underwent transthoracic and transesophageal echocardiography on the day before the intervention.A standard transthoracic echocardiogram was followed by a focused transesophageal echocardiogram (TEE),according to the standard operating procedures of our department.Either local pharyngeal anesthesia alone or in combination with conscious sedation using propofol or midazolam were used for the transesophageal examination.TEE included at least the following: (1) depiction of the left atrium appendage (LAA) in order to detect possible thrombi,(2) measurements of the LAA at various angles,and (3) biplane depiction of the interatrial septum with and without color doppler to detect atrial septal defects.LAA measurements were performed according to the instructions of the manufacturer of the occlusion device used.Nyquist (aliasing velocity) limit for color Doppler was set at 60 cm/s.

The LAAO procedure was performed under TEE guidance.LAA dimensions were once more measured in the catheterization laboratory,before choosing device size.A transthoracic echocardiography exam was performed 2 h after the intervention and prior to discharge from the hospital to detect possible device dislocation or pericardial effusion.

A follow-up TEE was performed at three months,according to the standard operating procedures of our department.The follow-up TEE was performed according to the same protocol used on the day before the procedure.

Percutaneous Left Atrial Appendage Occlusion

Percutaneous LAAO was performed in all patients by the same interventional cardiologist,with experience in the procedure.Device type and size were decided by the interventional cardiologist,after reviewing the pre-and periprocedural TEE measurements.Two types of LAAO devices were used,namely WATCHMAN FLX? (Boston Scientific)and Amplatzer? Amulet? (Abbott).Femoral venous access was obtained after local anesthesia and 5,000 international units of unfractionated heparin were administered.Conscious sedation with propofol infusion was used.Puncture of the fossa ovalis and implantation of the occlusion device were performed under TEE guidance using an SL1 transseptal sheath and a BRK XS transseptal needle (both from Abbott).An additional dose of 5,000 international units of heparin was administered after transseptal puncture.In patients with a patent foramen ovale (n=4),the latter was not used for passage and transseptal puncture was performed like in all other patients.After transseptal puncture,the transseptal sheath was advanced posteriorly into the left atrium and an Amplatz stiff guidewire was introduced in the left upper pulmonary vein.The SL1 sheath was removed and the LAAO introducer sheath (inner/outer diameter 12/14 Fr) was introduced over the guidewire into the left atrium.After removing the guidewire,the introducer sheath was placed in front of the LAA and the implantation device was advanced.After successful implantation,stable device position was confirmed by performing the so-called “tug test” (pulling of the occluding device) twice,before releasing the device.Intravenous heparin administration was repeated during procedures that lasted more than 30 min,aiming at an activated clotting time of >250 s.Following LAAO,patients received dual antiplatelet therapy with aspirin and clopidogrel for three months,followed by lifelong monotherapy with aspirin.After implantation,patients remained on the regular ward for at least two nights before discharge.

Hemodynamic Measurements

In order to assess the hemodynamic effects of transseptal puncture during LAAO,pressure measurements as well as measurements of oxygen saturation were performed in the left and right atrium before and after implantation of the occlusion device.The first measurement was performed in the right atrium before puncture of the fossa ovalis.Following transseptal puncture and advancement of the transseptal sheath,pressure and oxygen saturation were measured in the left atrium.After successful implantation of the occlusion device,a pigtail catheter was advanced through the introducer sheath into the left atrium and the introducer sheath was pulled back into the right atrium,before recording left atrial pressure through the tip of the pigtail catheter.This maneuver aimed at “opening” blood flow through the iASD,while allowing measurement of left atrial pressure.After a waiting period of 60 s,a blood sample for oxygen saturation analysis was obtained through the pigtail catheter.Finally,the pigtail catheter was pulled back into the right atrium and right atrial pressure and oxygen saturation were measured once more.Mean pressure recordings were used for the analysis.

In the cases where oxygen was delivered to the patient,pulse oximetry with a target oxygen saturation of 94% was used to guide oxygen therapy.Oxygen flow rate was kept stable during the whole procedure.

Statistical Analysis

Quantitative data are presented as mean ± SD and categorical data as frequency and percentage.Normal distribution was tested using D’Agostino &Pearson test.Comparison analyses between measurements in the same patients were performed using the pairedt-test for normally distributed data and the Wilcoxon matched-pairs signed rank test for data that was not normally distributed.Tests were two-tailed.Differences were considered statistically significant ifP<0.05.Comparison analyses and descriptive statistics were performed using GraphPad Prism version 8.0.

RESULTS

Study Population

Percutaneous LAAO was performed in a total of 53 patients.Five patients had an atrial septal defect and were excluded from the study.The study population finally consisted of 48 patients (38 males,10 females;mean age,79.6 ± 7.6 years).One patient was excluded from the analysis of the hemodynamic and follow-up parameters,because puncture of the fossa ovalis and device implantation was not possible.Baseline characteristics of included patients are summarized in Table 1.

LAAO Procedure

Implantation of the occlusion device was successful in 47 patients.Puncture of the fossa ovalis was not possible in one patient.An Amplatzer? Amulet? occlusion system was used in most cases (61.7%Amplatzer? Amulet?vs.38.3% WATCHMAN FLX?,Table 2).Periprocedural TEE using color doppler revealed an iASD in all forty-seven patients immediately after device implantation and removal of the introducer sheath.In most patients (41/47 patients,85.4%) a left-to-right shunt was revealed.The rest of the patients had a bidirectional shunt.No patients had constant right-to-left shunt.Periprocedural details are summarized in Table 2 and periprocedural complications in Table 3.

Table 1 Baseline characteristics of the study population.

Table 2 Periprocedural details.

Hemodynamic Measurements

In order to assess the immediate hemodynamic effects of transseptal puncture and iASD after percutaneous LAAO,mean pressure and oxygen saturation were recorded in both the right and the left atrium,as described in theMethodssection.No measurements could be performed in two patients,for reasons not related to the study.Moreover,oxygen saturation measurements were not complete in two more patients.

Measurements in the right atrium revealed a statistically significant increase of mean pressure after removing the introducer sheath (9.8 ± 5.5 mmHg before transseptal puncturevs.11.6 ± 5.0 mmHg after sheath removal,n=45,P=0.034;Figure 1A),whereas no significant difference was found in oxygen saturation (66.5% ± 8.3% before transseptal puncturevs.65.8% ± 10.0 % after sheath removal,n=43,P=0.623;Figure 1C).Pressure measurement in the left atrium showed no significant difference between mean pressure recordings directly after transseptal puncture and after removing the introducer sheath (19.2 ±9.0 mmHg after puncturevs.18.0 ± 8.1 mmHg after sheath removal,n=45,P=0.718;Figure 1B).No significant difference was found in oxygen saturation in the left atrium after removing the introducer sheath (97.8% ± 2.0% after puncturevs.97.2% ± 2.4%after sheath removal,n=45,P=0.129;Figure 1D).

Figure 1 Immediate hemodynamic effects of transseptal puncture.Graphical representation of mean pressure in the right(A) and left atrium (B),as well as oxygen saturation in the right(C) and left atrium (D) before and after opening iASD and allowing free blood flow through it.Lines represent mean values.Floating bars represent minimum and maximum values.＊P <0.05;iASD: iatrogenic atrial septal defects;ns: non-significant.

Persistence of iASD at Mid-term Follow-up

Follow-up TEE at three months revealed a per-sistent iASD in four out of forty-seven patients (8.5%).Periprocedural details,biochemical,echocardiographic,and hemodynamic parameters of these four patients are shown in Table 4.

Table 3 Periprocedural complications

Total procedure duration in these four patients with persistent iASD was not higher compared to the other individuals.Moreover,none of the patients having undergone relatively long procedures(duration above the mean duration of 57.4 min) had a persistent iASD at three months.Thus,no correlation appears to exist between the length of the interventional procedure and the persistence of iASD after three months.After reviewing the characteristics of patients with persistent iASD (Table 4),there appears to be no obvious correlation with occlusion device used,device size,presence of severe mitral or tricuspid regurgitation,and shunt direction after the procedure.Interestingly,two out of three patients on dialysis included in our study had a persistent iASD at three months.Moreover,three out of four patients with persistent iASD had an above-average mean right and left atrium pressure before implantation.

Table 4 Characteristics of patients with persistent iASD at 3 months.

DISCUSSION

The main finding of our study is that iASD after percutaneous LAAO does not result in significant shunts directly after the procedure,as no significant differences occur in oxygen saturation of the left and right atrium after withdrawing the transseptal sheath.However,iASD resulted in a significant increase of mean right atrial pressure.Persistent iASDs after percutaneous LAAO seem to be relatively rare at three months,with an incidence of 8.5%.

In 2003,a case-series with nine patients undergoing percutaneous LAAO with the PLAATO device was published by Omran,et al.[16]Only one of the patients (11.1%) had a post-procedural iASD that closed at three months.In our study,all patients had an iASD directly after the procedure,as demonstrated by periprocedural TEE with color doppler.However,the exact timepoint of iASD depiction was not reported in the study of Omran and coworkers (directly after implantationvs.before patientdischarge).This does not allow a direct comparison of our results with those of this case-series.The only relatively large study to date is the one by Signet al.,[14]in which 253 patients having undergone WATCHMAN implantation as part of the PROTECT AF study were included.Follow-up with TEE using echo-contrast was performed after the procedure,at 45 days,at 6 months and 12 months.Most patients (87%)had an iASD immediately after the procedure,whereas in our study an iASD could be found in all patients after the procedure.In agreement with our study,most patients had a left-to-right shunt (90%)after the procedure and there was a small number of patients with bidirectional (5%) or right-to-left shunt (4%).The incidence of iASD at follow-up in the study of Sign,et al.[14]was higher (34% at 45 days,11% at 6 months,7% at 12 months) compared to our study (8.5% at three months).Moreover,there were no patients with right-to-left shunt after the procedure in our study.Recently,Puga,et al.[15]published the results of a single-center,retrospective study performed in 70 patients undergoing percutaneous LAAO.The incidence of iASD at one month was 37% and was not associated with adverse clinical outcomes.In this study,procedures were guided using intracardiac echocardiography in 30 out of 70 patients.Intracardiac echocardiography was found to be an independent predictor of iASD at one month.In another recent study by Nelles,et al.[17]including 144 patients undergoing LAAO,approximately 35% of patients undergoing percutaneous LAAO had a persistent iASD at three months.The results of the above studies cannot be directly compared to our study,because of differences in study design and follow-up timepoints.However,it is apparent that a difference in the prevalence does exist,in particular when comparing our data with the work of Nelles,et al.(35%vs.8.5% in our study at 3 months).[17]The reasons for this discrepancy are not clear.As commented by Nelles,et al.,[17]factors such as differences in mean age (higher mean age in our study;79.6 ± 7.6 in our studyvs.75 ± 7 years in the study of Nelles,et al.[17]),sex (79.2% male in our studyvs.63.9%),and comorbidities (CHA2DS2-VASc 4.8 ± 1.3 in our studyvs.4.2 ± 1.2) may play a role in the observed differences between published studies.One would probably expect a higher prevalence of persistent iASD in an older patient population with more comorbidities.However,the prevalence of persistent iASD in our study is much lower.Variations in the interventional procedure and specifically in the manipulations during transseptal puncture and advancing of the transseptal sheath may also strongly influence the possibility of an iASD persisting in the mid-or long-term.More studies are needed to clarify the way all these parameters may influence the pathophysiology of iASDs.

Most published studies regarding iASD after puncture of the fossa ovalis have been conducted in patients with severe mitral regurgitation undergoing transcatheter edge-to-edge repair of the mitral valve using the MitraClip system.As already mentioned,guiding sheaths used in this procedure are larger,namely 24 Fr.Incidence rates of persistent iASDs reported in these studies are significantly higher.A study conducted in 30 patients undergoing mitral valve repair using MitraClip found an incidence of 27% at 12 months.[11]Patients with a persistent iASD had more residual mitral regurgitation,and increased tricuspid regurgitation.Noteworthy,iASDs at follow-up were identified using transthoracic echocardiography and no TEE was performed,meaning that the reported incidence may be underestimated.Another study with follow-up TEE at six months reported an iASD incidence rate of 62%,with female gender and reduced left ventricular ejection fraction being independent predictors of iASD persistence.[18]A long-term follow-up after at least 12 months using TEE was performed in the study of Paukovitsch,et al.[19]The authors reported persistent iASD in 41.7% of patients,all with left-to-right shunt,with no significant differences in the reduction of mitral regurgitation or in clinical parameters,such as dyspnoea,distance in 6-min walking test,mortality or rehospitalisation rates.Interestingly,fluoroscopy and device times (time between insertion of the guide catheter and its removal) were significantly longer in patients with a persistent iASD,thus supporting “the concept of manipulation at the atrial septum as a key mechanism leading to persisting iASD”.[19]Other studies also seem to support these results.[10]A recently published meta-analysis of six studies and a total of 361 patients reported a 1-year incidence of 28 %.[20]Pre-existing atrial fibrillation,residual mitral regurgitation,and prolonged fluoroscopic time were found to be predictors of an iASD.Moreover,patients with an iASD had a higher risk for rehospitalization.However,interventional closure of iASD after Mitraclip has not been shown to result in improved clinical outcomes compared to conservative therapy.[21]Another study has reported increased left atrial pressure after MitraClip procedure to be associated with persistent iASD at one and 12 months.[22]Because of the small number of patients with a persistent iASD in our study,no correlation analysis or statistical comparison could be performed,although total procedure duration in the four patients with persistent iASD was not increased (none of the four patients with persistent iASD had an above-average procedure duration).Noteworthy,three out of four patients with persistent iASD in our study had an above-average mean right and left atrium pressure before implantation.

Patients undergoing pulmonary vein isolation or other types of ablation therapy for atrial fibrillation represent another patient collective of interest,because transseptal puncture is required to allow advancement of the ablation catheter into the left atrium.Smaller transseptal sheaths of 8 or 8.5 Fr are mostly used in such procedures.Two transseptal sheaths are used by some electrophysiologists.Hammerstingl,et al.[12]published a study conducted in 42 patients undergoing pulmonary vein isolation using radiofrequency ablation.The interatrial septum was punctured once in 27 and twice in 15 of the patients using 8 Fr transseptal sheaths.Follow-up was performed using TEE at 9 months.Incidence of persistent iASD at follow-up was about 30% in patients with a single transseptal puncture whereas no iASD was found in the patient group with double transseptal puncture.These results suggested that double transseptal puncture with passage of 2 electrophysiological catheters may be less traumatic than passage of 2 electrophysiological catheters via a single puncture site.In another study by Sieraet al.,[23]incidence of iASD was assessed in 39 patients undergoing pulmonary vein isolation using cryoballoon ablation technique,using a single transseptal puncture and a 15 Fr catheter.A persistent iASD was observed in 20% of the patients at 12 months.An even higher incidence rate of 27% after cryoballoon ablation at a mean follow-up time of 2.9 years was reported by Linhart and co-authors,with persistent iASD not being associated with clinical complication or deterioration of echocardiographic parameters.[13]

To the best of our knowledge,this is the first study investigating the hemodynamics after transseptal puncture for LAAO by measuring pressures and oxygen saturation in the left and right atrium before and after transseptal puncture.Published evidence about hemodynamics mostly focuses on the persistence of mitral regurgitation after percutaneous edge-to-edge therapy.For example,more severe mitral regurgitation was found in the group with persistent iASD at 12 months in the study of Smith and co-authors.[11]In this study,more residual regurgitation may have been the cause of an increased left atrial pressure and,therefore,iASD persistence.Interestingly,there was no significant difference in left atrial volume between patients with and without a persistent iASD.In the recent study of Paukovitsch,et al.[19]there was no difference in the severity of residual mitral regurgitation between patients with or without a persistent iASD.There was,however,a significant difference in the severity of tricuspid regurgitation (more severe regurgitation in patients with an iASD).In another recent study in 69 patients undergoing a MitraClip procedure,the resulting iASD after removal of the MitraClip guiding catheter from the left atrium led to a left-to-right shunt with a mean Qp/Qs ratio of 1.09.[24]Right ventricular dimensions and functional parameters remained unchanged at 6 months follow-up.Patients with higher Qp/Qs ratios in this collective were not found to have significant differences in clinical and echocardiographic parameters compared to patients with lower shunt fractions.

In our study,an iASD could be observed in all patients after removal of the introducer sheath from the left atrium.A left-to-right shunt could be observed in most patients (85.4%),and a bidirectional shunt in the rest (14.6%).There were no patients with a right-to-left shunt.The iASD resulted in a significant increase in right atrial pressure.Interestingly,no significant reduction could be observed in left atrial pressure.The exact pathophysiological mechanisms leading to this latter,single-sided observation,are not clear.Lower pressures in the right heart may be one of the reasons leading to the pressure difference reaching statistical significance only in the right atrium.Oxygen saturation in the left and right atrium was not found to be different after removal of the introducer sheath.Again,one would expect the increase in right atrial pressure to be accompanied by an increase in oxygen saturation,due to oxygenated blood flowing from the left to the right atrium.However,this was not the case in our study,the reason being,once more,not clear.We cannot exclude,that factors other than left-to-right shunt may have contributed to the observed increase in right atrial pressure at the end of the procedure.In general,given the proven clinical benefit of interventional procedures requiring transseptal puncture,including percutaneous LAAO,we would not have expected hemodynamically significant iASDs to be the rule after transseptal puncture,because,in such a case,this benefit would have been mitigated by the significant interatrial shunt.The results of our study confirmed these expectations.

Our study has some important limitations.It is a relatively small study with a short follow-up time of three months.Due to the low number of patients with a persistent iASD,no statistical comparisons could be performed between patients with and without a persistent iASD.Correlation analyses that could reveal parameters associated with iASD persistence were not possible.Our patient collective undergoing percutaneous LAAO is not homogenous,as patients with varying clinical history and baseline echocardiographic parameters were included.Although our patients were prospectively included in the study and the hemodynamic measurements were performed as part of the study,our study protocol did not influence the interventional procedure or its preparation in any other way (other than the measurement of pressure and oxygen saturation).This makes it impossible to control factors such as procedure duration,depth of sedation,and volume status.Procedure duration may strongly influence patient hemodynamics and this may result in altered pressures in the left or right atrium.A longer procedure may also result in stronger forces/ manipulations applied to the interatrial septum by the transseptal sheath,which may in turn influence the possibility of the iASD persisting in the mid-or long-term.Conscious sedation was achieved using propofol infusion.Propofol is known to have direct and indirect vasodilatory effects,thus affecting arterial pressure and,as a result,intracardial pressures.We cannot exclude that differences in propofol dose or in the physiological response of different patients’ circulatory system to propofol infusion may have influenced our results.Additionally,differences in volume status caused by comorbidities such as chronic heart failure or diuretic therapy may also have affected invasive pressure measurements and shunt directions.Finally,other parameters,such us mitral or tricuspid regurgitation,size and distensibility of the atria,heart rate and rhythm (sinus rhythmvs.atrial fibrillation) may have affected our measurements.

In conclusion,transseptal puncture during percutaneous LAAO results in an iASD after removal of the transseptal sheath.The resulting shunt leads to a significant increase of right atrial pressure.Left atrial pressure and oxygen saturation in the left and right heart are not significantly influenced.Persistent iASD at mid-term are observed in 8.5% of patients.Larger studies are needed in order to unravel the pathophysiology and hemodynamic results of iASD after percutaneous LAAO.