Zoltan G. Turi

Iatrogenic Pericardial Effusion and Tamponade in the Percutaneous Intracardiac Intervention Era

The number, specific type, and complexity of percutaneous intracardiac procedures continue to evolve. Many of these procedures require left atrial access using transseptal techniques. These approaches carry with them the potential for pericardial effusion (PE) and cardiac tamponade, particularly in the setting when intraprocedural anticoagulation is being administered. PEs and even cardiac tamponade have been documented with both diagnostic as well as therapeutic procedures. When the effusion is a complication of an intracardiac procedure, it is usually the result of a cardiac perforation. The presentation depends on several factors including the structure that is perforated, the device that caused the perforation, the baseline hemodynamic status of the patient, and the level of anticoagulation present. The incidence has varied substantially although it has been recorded with essentially all intracardiac procedures, both diagnostic and therapeutic on both the right and left side of the heart. Prompt recognition is essential so that prevention of the transition from effusion to tamponade can be attempted (e.g., by reversing anticoagulation) or the hemodynamic collapse can either be averted or treated. Clinical, radiologic, and echocardiographic assessment are each important. Pericardiocentesis can be life-saving and is a core competency for all laboratories performing invasive cardiac procedures. Systems of care must include the knowledge base, equipment, and expert echocardiographic and interventional personnel. Collaboration with noninvasive colleagues and training interventionalists who perform intracardiac interventions, both electrophysiologists and interventional cardiologists, should be required as part of every invasive program.

Optimizing vascular access: Routine femoral angiography keeps the vascular complication away

The study by Sherev et al. in this issue represents a minor landmark in the very gradual half-century evolution of the Seldinger technique. While their findings that retroperitoneal bleeds occurred with high (or near-high) sticks and that hematomas were more common when femoral artery puncture missed the area over the femoral head were not surprising, their speculation regarding the potential relevance of the course of the inferior epigastric artery is thought-provoking if not proven by the data presented. There has been a long debate in these pages regarding the merits of femoral artery angiography prior to deployment of vascular closure devices [1,2]. This has been quite remarkable, because despite the admonition in the instructions for use for vascular closure devices requiring routine predeployment angiography, an amazingly large number of interventionalists have resisted this step, arguing in the face of a solid evidence base that it is unnecessary. In contrast, we have suggested repeatedly that femoral angiography should be performed even when closure is by manual compression [3]. The authors in the accompanying article have provided an argument not only for routine femoral angiography, but also for performing it at the very beginning of the catheterization. There are a number of methodological issues in the paper by Sherev et al. that should be pointed out because they potentially muddle the conclusions. The control group was randomly selected rather than a matched cohort, resulting in baseline differences shown in their Table I. There is a significant possibility of type II error in the between-group comparisons. The complications for which meaningful statistical analyses are done are largely hematomas, which are rarely life-threatening, so that the authors’ conclusion that access site predicts life-threatening complications is rather broad, especially since life-threatening complications such as limb ischemia and sepsis are not included. The occurrence of six retroperitoneal bleeds, all in their groups 3 and 4, is suggestive and interesting, but larger numbers will be needed to confirm their speculation, which is based on very small subgroup comparisons that are not statistically significant. Moreover, it would be a mistake to assume complacently that retroperitoneal bleeding would be limited to this subset, as the authors have also pointed out. The authors’ use of a 10 cm threshold for hematoma deserves comment. In this journal alone, this reviewer has noted definitions of 2, 4, 5, 7, 10, and 15 cm by various studies. The inverse correlation between threshold size and incidence of hematomas is intuitively obvious and further confounds attempts at metaanalyses of access siteand closure-related complications [4]. The authors’ multivariate analysis did not include a number of variables suspected or known to affect complication rates, including sheath size, heparin dosage and ACTs, use of glycoprotein IIb/IIIa inhibitors, presence of atherosclerotic disease in the femoral artery, and use of vascular closure devices (although the latter three were included in the univariate analyses). There have been prior data suggestive of vascular device-associated increase in post-PCI retroperitoneal bleeding up to 10-fold, particularly in the setting of glycoprotein IIb/IIIa use [5]. Because prior studies did not report preclosure device femoral angiography (and in fact angiograms may not have been performed in most of the patients whose results are reported in the literature), it is unknown whether high sticks may have been an associated factor. The higher complication rate with manual and mechanical compression noted by Sherev et al. may well reflect strong selection bias in this retrospective study, including the possibility of a Hawthorne effect. The lack of significance of body surface area, diabetes, female gender, and other variables previously shown repeatedly to impact complication rates also strongly suggests the effect of these same limitations on the data base and analyses.

Clinical Trial Principles and Endpoint Definitions for Paravalvular Leaks in Surgical Prosthesis: An Expert Statement

The VARC (Valve Academic Research Consortium) for transcatheter aortic valve replacement set the standard for selecting appropriate clinical endpoints reflecting safety and effectiveness of transcatheter devices, and defining single and composite clinical endpoints for clinical trials. No such standardization exists for circumferentially sutured surgical valve paravalvular leak (PVL) closure. This document seeks to provide core principles, appropriate clinical endpoints, and endpoint definitions to be used in clinical trials of PVL closure devices. The PVL Academic Research Consortium met to review evidence and make recommendations for assessment of disease severity, data collection, and updated endpoint definitions. A 5-class grading scheme to evaluate PVL was developed in concordance withxa0VARC recommendations. Unresolved issues in the field are outlined. The current PVL Academic Research Consortiumxa0provides recommendations for assessment of diseasexa0severity, data collection, and endpoint definitions. Future research in the field is warranted.

Concomitant coronary and multiple arch vessel stenoses in patients treated with external beam radiation: Pathophysiological basis and endovascular treatment

External beam radiation‐induced stenoses isolated to the coronary arteries or peripheral vessels have been previously described. We report for the first time the clinical presentation of two patients with concomitant coronary artery and multiple arch vessel disease following external beam radiation of the chest. We review the pathophysiology, discuss the treatment options and describe the percutaneous treatment of coronary, carotid, subclavian, and axillary stenoses related to this rare but likely underdiagnosed disorder. Catheter Cardiovasc Interv 2004;62:385–390.

Whom do you trust? Misguided faith in the catheter‐ or Doppler‐derived aortic valve gradient

The article by Sakthi et al. in this issue raises troubling questions regarding the current practice of evaluation of aortic stenosis. The authors emphasize the possibility that Doppler-derived peak gradients may overestimate the severity of stenosis. It is important to point out that catheterization-derived gradients may also underestimate the stenosis severity. The invasive measurement of aortic valve hemodynamics has evolved only modestly in the decades since Gorlin and Gorlin [1] first applied the principles of hydraulics to cardiac valve area measurement. Catheter-derived gradients, best obtained through simultaneous micromanometer recording on either side of the aortic valve, are now rarely obtained in that fashion. Most institutions approximate a gradient through an aortic valve pullback, assuming that beat-to-beat variation is minimal. The potential for technical errors using this method is substantial, in part because of the effect of respiratory variation. The accompanying article uses another method, simultaneous measurement of left ventricular and systemic pressure through a long fluid-filled catheter in the left ventricle and a fluid column in the side arm of a sheath in the external iliac artery placed through the common femoral approach. The potential error using this commonly applied technique is substantial, can underestimate the peak and variably underor overestimate the mean gradients significantly [2], and has been described as ‘‘the least desirable approach’’ because of the effect of arterial impedance and the role of harmonics on the femoral as compared to the central aortic pressure [3]. Thus, the likelihood is that the catheter derived aortic valve gradients reported by Sakthi et al. are low at least in part because of the errors introduced by the method of measurement. Although the authors state that ‘‘close correspondence of the ascending aortic and femoral artery pressures was confirmed in all patients,’’ the typical variation in peak gradient is in the range of 10–20 mm Hg and would explain a number of the data points in the accompanying article. The 40 mm Hg difference between catheterand Doppler-derived gradients in two of the patients is difficult to reconcile regardless of the technique. It is more likely partly explained by a second major source of error: the fact that gradient is highly sensitive to preload, heart rate, catecholamine state, cardiac output, and a variety of other parameters, and that the measurements took place days or weeks apart. Not only were the cited variables likely substantially different, but differences in actual valve area and in left ventricular function could also have muddled the findings. Finally, use of peak-to-peak gradients, common in many laboratories, would have further underestimated the severity of aortic stenosis as described above [3]. There are important subtleties of difference between peak gradient, peak-to-peak gradient (which can substantially underestimate the peak), and mean gradient. The aortic valve is generally best assessed by valve area rather than gradient, although a case has been made for using a Gorlin constant independent measure such as valve resistance [4]. Whether any or all of the above could explain the 0.8 6 0.3 cm aortic valve area by echo vs. the 2.0 6 0.9 cm by catheterization would require an understanding of important data points not included in the authors’ Table I, including the parameters cited above. A comparison of tracings in Figure 1 demonstrates the variability introduced by the measurement technique used. The ideal technique for invasive measurement is simultaneous recording of pressures on either side of the aortic valve. This can be accomplished by several means: two arterial catheters, one in the LV and one in the central aorta; one dual-lumen catheter placed across the aortic valve; a diagnostic catheter placed in the central aorta with a pressure wire placed across the aortic valve; or a catheter placed via transseptal puncture into the left ventricle with a second catheter placed retrograde into the central aorta. Each technique has limitations, either because of enhanced risk, limited availability of the technology, or potential artifact induced by the measurement technique itself. The first

Complications of vascular closure devices—not yet evidence based

The interesting publication by Dangas et al. [(1)][1]claims to compare arteriotomy closure devices with manual compression after percutaneous coronary intervention. Unfortunately, this is a comparison only in the fashion that can be ascribed to a retrospective trial with mismatched procedural

Clinical trial principles and endpoint definitions for paravalvular leaks in surgical prosthesis

The VARC (Valve Academic Research Consortium) for transcatheter aortic valve replacement set the standard for selecting appropriate clinical endpoints reflecting safety and effectiveness of transcatheter devices, and defining single and composite clinical endpoints for clinical trials. No such standardization exists for circumferentially sutured surgical valve paravalvular leak (PVL) closure. This document seeks to provide core principles, appropriate clinical endpoints, and endpoint definitions to be used in clinical trials of PVL closure devices. The PVL Academic Research Consortium met to review evidence and make recommendations for assessment of disease severity, data collection, and updated endpoint definitions. A 5-class grading scheme to evaluate PVL was developed in concordance with VARC recommendations. Unresolved issues in the field are outlined. The current PVL Academic Research Consortium provides recommendations for assessment of diseasexa0severity, data collection, and endpoint definitions. Future research in the field is warranted.

A multicenter randomized trial comparing the effectiveness and safety of a novel vascular closure device to manual compression in anticoagulated patients undergoing percutaneous transfemoral procedures: The Celt ACD trial

This study compared the performance of Celt ACD®, a novel stainless steel based vascular closure device versus manual compression (MC) for femoral arteriotomy site hemostasis in patients undergoing percutaneous coronary procedures.

Taking a bite out of the femoral artery: vascular closure and vascular obstruction.

Trauma to the femoral artery, both from access as well as closure, has been recognized since Seldinger introduced the percutaneous technique five decades ago. The dramatic expansion of endovascular procedures has increased the proportion of the population undergoing vascular puncture, and in particular is increasing the number of patients whose arteries are repeatedly instrumented. Leaving the arterial lumen intact is likely to be of ever increasing import. The article on claudication secondary to suture-mediated vascular closure in this issue of the Journal describes nine cases in which the authors feel that vascular insufficiency was caused by sutures placed at the femoral artery puncture site. While I believe the premise that vascular closure devices can be responsible for varying degrees of vascular insufficiency is correct, a number of elements that would help the reader understand the relevance of these interesting cases is missing. First, it would be helpful to know the denominator: were these 9 cases culled from a single institution experience and how many total patients had the Perclose device placed? Since the number of cases of what the authors believe is suture related obstruction is so high, the authors appear to imply that other institutions with similar case volumes may simply be missing the diagnosis. If so, what is happening to those patients? Or are there atypical features to the authors’ patient population or procedures that might help explain the high number of cases of vascular insufficiency? Finally, is it possible that some of these cases in fact do not represent device related vascular obstruction? It would have been helpful to know the pre-closure size of the femoral artery and whether there was preclosure stenosis. The authors are silent on the issue of pre-closure femoral angiography in their patients and the very fact that Perclose was deployed in some of these cases raises the question of whether pre-closure angiography was performed. It is puzzling that this step is still routinely skipped by many institutions and operators [1]. The current series of 9 cases includes device closure when sheaths were placed in the superficial femoral and external iliac arteries. The former is an off-label use that occasionally makes clinical sense: some operators close antegrade punctures after angioplasty of the SFA to avoid compression and transient diminished flow across a fresh angioplasty site. However, the SFA is frequently small, and the Perclose device is not designed to be used in arteries that are <5 mm. One would not want to knowingly deploy Perclose in the SFA at what appears to have been just below or at the level of the femoral bifurcation in patient number 9, again raising the issue of pre-closure angiography. Similarly, the external iliac closure in case number 4 is also problematic with two specific concerns coming to mind. One is the risk of retroperitoneal hemorrhage, the other that tethering of the vessel with sutures placed above the inguinal ligament will contribute to the appearance seen in Figure 4. In general, use of closure devices outside the common femoral artery, although touted for the titanium staple based device AngioLink (Medtronic, Minneapolis, MN) has little evidence base [2], and in my opinion is potentially dangerous, especially with high sticks in fully anticoagulated patients such as case number 4 [3]. The next issue to consider is the spectrum of leg pain that can be attributed to vascular closure devices. In my experience, vascular insufficiency at the femoral puncture site can cause a panoply of symptoms, including hypesthesias of the hip in addition to claudication. A common cause of leg pain after catheterization is hematoma, with compression of accessory branches of the femoral nerve causing both pain and numbness radiating to the knee being relatively common. Vascular closure devices have been associated with occasional nerve irritation, and pseudoaneurysms cause substantial pain with pulsatile expansion. Nerve entrapment has been anecdotally described, primarily with suture closure devices but also with several other technologies. There have been unpublished descriptions of surgery to remove closure devices to relieve pain caused by presumed stitch entrapment, but the evidence base has been weak. Surgery has a powerful placebo effect so that relief of symptoms alone would not

Does a vascular closure device affect the imbalance in complications between radial and femoral access

There exists a robust evidence base of superior outcomes in ST-Elevation Myocardial Infarction (STEMI) with radial as opposed to femoral access Benefit of vascular closure devices to decrease femoral access complications remains a matter of contention; the evidence base is largely incomplete Vascular site complications and access site related bleeding may be as much, or more affected by access technique as by method of vascular closure