Aortic Stenosis Catheterization Revisited: A Long Sheath Single-Puncture Technique

Aortic stenosis is one of the most common forms of valvular heart disease seen in adults. Despite advances in echocardiography, the evaluation of aortic stenosis continues to be a frequent procedure in the cardiac catheterization laboratory. In order to calculate the aortic valve area using the Gorlin formula,¹ the mean transvalvular pressure gradient across the aortic valve must be measured. To obtain this transvalvular gradient, simultaneous evaluation of the proximal aortic and left ventricular pressures yields the most accurate data. Classically, this has required placing one arterial catheter in the aortic root via one femoral artery, and the second catheter in the left ventricle via the other femoral artery. (Alternatively, one catheter can be placed in the left ventricle via the transseptal approach.) However, in addition to causing more patient discomfort and entailing more risk, these dual-puncture techniques can be difficult in patients with peripheral vascular disease. Therefore, several single-puncture techniques have been tried, all with significant disadvantages.^2–7 One successful single-puncture technique used was an 8 Fr double-lumen pigtail catheter. This provided very good transvalvular gradients,⁸ and was in clinical use for about 10 years. However, this catheter was withdrawn from the market in 2000, and at the time this study was performed, a good substitute was not readily available. Fusman^2,9 proposed using a pressure wire technique to measure the gradient, but this technique is not available to the noninterventionalist. A Millar double-lumen, high-fidelity catheter with a side hole fluid lumen was commercially available, but was not used in many catheterization laboratories due to the expense and technical expertise required. A 6 Fr double-lumen, fluid-filled catheter has recently been made available (Vascular Solutions, Inc., Minneapolis, Minnesota), but was not available at the time that this study was performed. In 1989, Nath et al. proposed using an 8 Fr 59-cm sheath (originally designed for transseptal use) with a 5 Fr pigtail catheter to obtain simultaneous left ventricular-aortic pressure gradients, with good results.¹⁰ This technique never was widely adopted, and since that time, increasing numbers of patients with aortic stenosis have had a prior CABG, including the use of a left internal mammary artery. Using the very long sheath does not allow access to the mammary artery, and may prevent adequate engagement of saphenous grafts as well. In recent years, carotid intervention has become more common, and with it the introduction of 55-cm long sheaths that stop just distal to the origin of the subclavian artery. Additionally, the size of all catheters and sheaths have decreased, still with good hemodynamic fidelity. This led to the hypothesis of the current study: that combining a 4 Fr pigtail catheter within a 6 Fr 55-cm long sheath would provide as accurate a measurement of the transvalvular gradient as can be obtained from the traditional two-puncture technique. Materials and Methods Patient selection. The study was approved by the institutional review board of the South Texas Veterans Health System, San Antonio, Texas, and informed consent was obtained from patients prior to the study. All patients referred to the Audie L. Murphy Memorial Veterans Hospital cardiac catheterization laboratory with a diagnosis of aortic stenosis were screened for eligibility. Patients were excluded if they had: 1) atrial fibrillation/flutter; 2) significant peripheral vascular disease precluding the use of 6 Fr catheters in both femoral arteries; 3) more than mild aortic or mitral regurgitation; 4) any other valvular stenosis. The 13 patients enrolled in the study (hereafter called the Test Population) all underwent a recent echocardiogram confirming significant aortic stenosis (defined as an aortic valve area less than or equal to 1.10 cm²), although some echos were performed at the outlying referring hospitals and were not repeated at this institution. (Average echo aortic valve area = 0.84 cm², range 0.56–1.10 cm²). During this same time period, other patients requiring an aortic stenosis catheterization who were not enrolled in this research protocol (either due to an exclusion, or the screener was not notified of the patient’s presence) had the same technique used via a single femoral approach at the discretion of the primary operator. The data from these cases (termed the Retrospective Population) were retrospectively reviewed. Methods – Cardiac Catheterization Three arterial transducers, and one venous transducer were used. A standard 6 Fr 10-cm arterial catheter was placed in one femoral artery, and a standard 8 Fr 10-cm venous catheter was placed in a femoral vein. A 55-cm long 6 Fr arterial catheter (Vista Brite Tip®, Cordis Corp., Miami, Florida; or RAABE, Cook, Inc., Bloomington, Indiana) was then placed in the second femoral artery, and the distal radiopaque tip was positioned under fluoroscopy just distal to the origin of the left subclavian artery. All the sheaths were flushed, followed by administration of 3,000 units of heparin through the side arm of the long sheath. The long sheath was then frequently flushed (every 5 to 10 minutes) during the case, and 1,000 units more of heparin were given at 60 minutes if the case was still in progress. A Swan-Ganz catheter was then advanced to the pulmonary artery. A 4 Fr pigtail catheter was passed up the long arterial sheath until the tip and all side holes exited the distal tip of the sheath, and a 6 Fr pigtail catheter was then passed up the short arterialsheath to the level of the other pigtail and long sheath tip. Pressures from the side arm of the long sheath and the two pigtail catheters were all then carefully zeroed, matched and recorded. Next, the 4 Fr pigtail was advanced to the aortic root just above the valve, and any pressure difference between it and the sheath tip was recorded. The 4 Fr pigtail was then advanced across the aortic valve, typically using a straight wire. Occasionally the valve was crossed first with an alternate 6 Fr catheter (such as a JR4) and then exchanged for the 4 Fr pigtail catheter. The 6 Fr pigtail catheter was then repositioned in the aortic root, and careful simultaneous pressure measurements were made from the left ventricle, aortic root and distal sheath tip (Figure 1). Using care to ensure a steady heart rate, pressures were recorded from the left ventricle-aortic root (control technique), and then from the left ventricle-distal sheath tip (new technique). Three to five sinus beats were recorded and the mean pressure gradient planimetered by the computer (GE Marquette Midas and GE Prucka 7000 systems, GE Healthcare Technologies, Waukesha, Wisconsin). Cardiac outputs were then obtained by thermodilution, assumed Fick and measured Fick methods. The aortic valve area and mean transvalvular gradient for at least 3 beats were recorded for each output, first from the left ventricle-aortic root, and then from the left ventricle-distal sheath tip. The Swan-Ganz was then removed, and a left ventriculogram was performed in the standard fashion using the 4 Fr pigtail catheter. The pigtail catheters and long arterial catheters were then flushed, and a careful pullback of the 4 Fr pigtail into the aortic root was recorded. Both pigtails were then pulled back to the level of the distal sheath tip, and a final match was recorded. The pigtail catheters were then removed. Coronary/graft angiography then proceeded in the usual fashion, using standard 6 Fr catheters through the long sheath and flushing the sheath after each catheter use. At the end of the case, the operator had the discretion of pulling the long sheath and using manual pressure, or exchanging the long sheath for a short 10-cm arterial sheath, performing a sheath angiogram, and then using a closure device if appropriate. The patient was followed for 24 hours after catheterization for any apparent procedure-related complications. Endpoints and Analysis Endpoints. The primary endpoint was to compare the difference in the mean pressure gradient between the left ventricle and the aortic root (control technique) versus that obtained between the left ventricle and the long sheath (new technique). Secondary endpoints were to compare the aortic valve area obtained from the control technique versus the new technique, and the difference in mean arterial pressure between the aortic root and the tip of the long sheath. In the retrospective population, the data were retrospectively analyzed for overall complication rate, length of procedure and procedural characteristics. Statistical analysis. Linear regression analysis was performed to compare transvalvular gradients and aortic valve areas obtained by the two techniques. The Student’s t-test was performed to determine if the correlation observed for the study sample was significantly greater than 0.75, if the slope of the regression line was significantly different from 1, and if the intercept of the regression line was significantly different from 0. SPSS, Inc. (Chicago, Illinois) and Stata Corporation (College Station, Texas) statistical analysis programs were used. Results Test population: Clinical characteristics. Thirteen male patients were enrolled in the study, ages 50–84 years (Table 1). The height of the patients ranged from 66–73 inches (170–185 cm). Two patients had undergone a prior CABG; 5 patients had no significant coronary disease. Of the 12 patients who underwent left ventriculography, the median left ventricular ejection fraction was 60% (range 51–82%). (One patient did not undergo left ventriculography due to an elevated left ventricular end-diastolic pressure). There were no complications noted in any of the patients. Transvalvular gradients and valve areas. Table 1 shows the systolic arterial pressure in the aortic root and at the tip of the long sheath. While there was often a slight difference in the pressures, there was no trend for one location being consistently higher than the other. There was no statistical difference in the mean arterial pressure between the two locations (r2 = 0.99, Figure 2). Therefore, the mean pressure gradient between the left ventricle and the tip of the long sheath was not adjusted for this difference in any patient. Table 1 also shows the mean transvalvular gradients and aortic valve areas obtained by the control method (with the 6 Fr pigtail in the aortic root) versus the long sheath method. As displayed in Figures 3 and 4, there was excellent correlation between the mean transvalvular gradients (r2 = 0.99), and aortic valve areas (r2 = 0.97). The median aortic valve area for the Test Population was 0.92 cm², with a range of 0.38–1.43 cm². Retrospective population results. Over 18 months, 7 different operators performed aortic stenosis catheterization on 55 other patients (ages 50–82 years, height 63–74 inches, 160–190 cm) using a single arterial puncture with use of the long 6 Fr sheath and a 4 Fr pigtail catheter. Of these 55 patients, significant coronary disease was found in 70%. The median aortic valve area was 0.90 cm² (range 0.47– 1.67 cm²). The median left ventricular ejection fraction in the 36 patients who underwent left ventriculography was 60% (range 26–81%). The aortic valve could not be crossed in 3 patients. The average time to complete a straightforward case without grafts or other angiography was 71 minutes ± 21 minutes. Two patients had anomalous coronary arteries which were engaged via the long sheath. Other procedures such as carotid or renal angiograms were also occasionally done during the same case without difficulty. In the cases of the renal/abdominal aortograms, the long sheath was pulled back to allow access to the abdominal aorta. Ten patients had undergone a prior CABG, 8 of which included a left internal mammary artery graft. In 4 of the graft cases, the operator chose to exchange the long sheath for a short sheath before engaging the coronary arteries or grafts. In the remaining graft cases, there was no difficulty engaging the grafts (including the left internal mammary grafts) via the long sheath. In the 45 nongraft cases, the long sheath was exchanged for a short 6 Fr sheath in 19 patients prior to engaging the coronary arteries. One patient had the long sheath placed through an aorto-femoral graft without any difficulty. A closure device was employed in 29 patients (20 Angio-Seal™ devices, St. Jude Medical, St. Paul, Minnesota devices; 9 Perclose devices, Abbott Laboratories, Redwood City, California), after exchanging the long sheath for a short one and performing a sheath angiogram. Complication rate: Both populations. In the two patient populations combined (n = 68), there was 1 cerebrovascular accident and 2 minor femoral bleeds without a hematocrit drop. Both patients with the femoral bleeds had received a Perclose closure device. The cerebrovascular accident occurred in a patient with known cerebrovascular disease who had originally been admitted for recurrent transient ischemic attacks. The patient had received 2,500 units of heparin prior to crossing the aortic valve, and another 1,000 units during subsequent graft engagement; the symptoms began after the LIMA engagement. The overall complication rate, therefore, was 1.5% for major complications, and 2.9% for minor complications. Discussion This study suggests that using the single-puncture long sheath technique is comparable to the gold standard of two arterial punctures, with excellent correlation of both the mean transvalvular gradient and aortic valve area. This correlation was excellent over a wide range of gradients, and valve areas (16–64 mmHg and 0.4–1.4 cm², respectively). Although the long sheath tip is several centimeters away from the proximal aortic root, in most cases there was no significant difference seen in the mean arterial pressure between the two locations, and a mean gradient could not actually be planimetered by the catheterization laboratory computer. This was true despite the variety of heights of our patients. Figure 5A (patient 7) shows the typically observed closely aligned pressure waveforms obtained from the left ventricle, aortic root catheter and the long sheath. Figure 5B shows the worst case of waveform alignment of the 13 patients. In this case (patient 5), the time delay between the left ventricle and the aortic root was 16 msec, versus a time delay of 29 msec between the left ventricle and long sheath; yet the unadjusted mean gradients were not different (49.0 versus 49.1 mmHg). Figure 5C shows the waveform differences in patient 6, who had the largest systolic blood pressure discrepancy between the aortic root and the long sheath. The unadjusted mean gradients of 49.9 mmHg and 53.1 mmHg in this case were clinically consistent, and the aortic valve areas were not significantly different (0.47 cm² versus 0.48 cm²). Although some authors have raised a concern about the pressure recovery phenomenon with the diminution of the pressure gradient as the catheter is pulled back higher in the aortic root,^11–13 this did not appear to be clinically significant in our study. This may be because pressure recovery has been suggested to only be of significance in small aortas,^14,15 which was not the case in our male patients. Nonetheless, any potential gradient difference between the aortic root and the tip of the long sheath should be specifically evaluated prior to crossing the aortic valve. Despite some initial concerns that the 4 Fr pigtail catheter may have dampened waveforms compared to the 6 Fr catheter, there was no difference seen in the pressure waveform of either the 4 Fr pigtail, the 6 Fr pigtail or the 6-Fr long sheath (Figure 5). In 2 patients, the 4 Fr pigtail was actually exchanged for a 6 Fr pigtail while in the left ventricle, and there again was no difference in the waveforms. Therefore, there was not felt to be any error from the different catheter sizes. Although the test population only consisted of 13 patients, the data were consistent over a wide range of aortic valve areas. The median aortic valve area and left ventricular ejection fraction for the test population were very similar to that of the larger retrospective population. This population of 55 patients studied with the single-puncture long sheath technique represents the cross section of typical patients with aortic stenosis presenting to a Veterans Hospital cardiac catheterization laboratory, many of whom have advanced vascular disease. Several operators adopted the technique, and were able to perform a variety of procedures during the same catheterization. Most operators found that while the 4 Fr pigtail catheter was too floppy to pass directly across the stenotic valve in a majority of cases, once a wire was across the valve, it was significantly easier to then pass the 4 Fr catheter over the wire compared to other catheters. In 3 patients out of the two groups combined, the valve could not be crossed (3/68, or 4%). This failure rate is typical of the spectrum of aortic stenosis cases, where the opening may be very eccentric. Most authors do not report their failure-to-cross rates, although Omran reported a 2% rate in his study.¹⁶ Although in about 40% of the cases the operator exchanged the long sheath for a short sheath prior to engaging the coronaries or grafts, in the remaining cases, the long sheath did not hamper any graft engagement. Many of the sheaths were then exchanged for a short sheath prior to engaging the coronary arteries, both to facilitate a sheath angiogram at the end of the case, and due to potential concerns about thrombosis. A potential complication of using a long sheath is thrombosis;⁹ for this reason, the sheath was aspirated and flushed frequently, and all patients received heparin. Although no complications were observed in the test population, one CVA was seen in the larger retrospective population (in a patient with known cerebrovascular disease). It should be noted that aortic stenosis catheterizations carry a higher risk of both silent (22%) and clinically apparent (3%) cerebrovascular events than the average diagnostic catheterization, according to Omran.¹⁶ In Omran’s series, this event rate was seen despite the use of 5,000 units of heparin. Therefore, use of the long sheath did not seem to specifically contribute to this complication, as our complication rate did not exceed this. Nonetheless, it became a frequent practice with most operators as time went on to exchange for a standard 10 cm sheath after the left ventriculogram, in addition to the heparin and frequent sheath flushing. The ability to exchange a short sheath for a sheath angiogram and subsequent closure device was popular with our operators, both in time saved and patient comfort. Despite the sheath exchanges, no increase in groin complications was noted. Study limitations. The main limitation of our study is that only male patients taller than 63 inches (166 cm) were evaluated, either prospectively or retrospectively. We found that in patients shorter than 66 inches (185 cm), the sheath often could not be advanced fully within the body if we wished to remain distal to the subclavian origin. In practice, however, the sheath was usually inserted all the way, and then just pulled back if needed to access the internal mammary artery. In theory, shorter men or women may do just as well with this technique, as long as the ascending aorta is > 3 cm wide to obviate concerns about pressure recovery in the smaller aortas. This needs to be validated, however, as does the technique in very tall patients. In our experience, in patients taller than 70 inches (178 cm), the sheath tip was several centimeters distal to the subclavian origin, but this did not cause a significant change in the mean arterial pressure in the heights studied (up to 74 inches [188 cm]). Nonetheless, in very tall patients, this potential problem needs to be considered. Another potential limitation for this technique is the possible loss of fidelity or leak through the valve sheath, which is true for any fluid-filled system. This was not appreciated in our patients, but as in any hemodynamic case, the operator must be on guard against over- or underdamped waveforms and re-zero the waveforms on a periodic basis. Conclusions In conclusion, we have shown that using a commercially available 55-cm long 6 Fr arterial sheath in combination with a 4 Fr pigtail catheter provides accurate evaluation of the mean transvalvular gradient and aortic valve area in the cardiac catheterization laboratory. This can be performed using a single arterial puncture and still allow for evaluation of grafts and other procedures as indicated. This technique is particularly useful in a busy catheterization laboratory where it is difficult to realign pressure gradients obtained by other methods, or where many patients have peripheral vascular disease precluding two arterial punctures. It also obviates the need to keep special dual-lumen catheters in stock just for aortic stenosis cases, as both the long sheath and the 4 Fr pigtail catheter have other uses in the catheterization laboratory as well.

1. Gorlin R, Gorlin G. Hydraulic Formula for calculation of area of stenotic mitral valve, other cardiac valves and central circulatory shunts. Am Heart J 1951;41:1–29. 2. Folland ED, Parisi AF, Carbone C. Is peripheral arterial pressure a satisfactory substitute for ascending aortic pressure when measuring aortic valve gradients? J Am Coll Cardiol 1984;4:1207–1212. 3. Fusman B, Faxon D, Feldman T. Hemodynamic rounds: Transvalvular pressure gradient measurement. Catheter Cardiovasc Interv 2001;53:553–561. 4. Folland ED, Parisi AF, Comei C. Simplified method for estimating true aortic valve mean gradient from simultaneous left ventricular and peripheral arterial pressure recordings. Cathet Cardiovasc Diagn 1990;20:271–275. 5. DeGeratu FT, Kreulen TH. Calculation of the aortic valve gradient from the left ventricular-femoral and aortic-femoral pairs of pressure waveforms. Cathet Cardiovasc Diagn 1998;44:9–13. 6. Kreuger SK, Orme EC, King CS, Barry WH. Accurate determination of the transaortic valve gradient using simultaneous left ventricular and femoral artery pressures. Cathet Cardiovasc Diagn 1989;16:202–206. 7. Brogan WC, Lange RA, Hillis LD. Accuracy of various methods of measuring the transvalvular pressure gradient in aortic stenosis. Am Heart J 1992;123:948–953. 8. Jayne JE, Catherwood E, Niles NW, Friedman BJ. Double-lumen catheter assessment of aortic stenosis: Comparison with separate catheter technique. Cathet Cardiovasc Diagn 1993;29:157–160. 9. Fusman B, Jolly N, Feldman T. Use of pressure wire to measure gradients across aortic valve. Am J Cardiol 2000;86(Suppl I):68I. 10. Nath A, Vetrovec GW, Mukharji J, et al. Simultaneous left ventricular and ascending aortic pressure measurements via single artery access for assessment of aortic stenosis. Cathet Cardiovasc Diagn 1989;17:126–130. 11. Levine RA, Cape EG, Yoganathan AP. Pressure recovery distal to stenoses: Expanding clinical applications of engineering principles (editorial). J Am Coll Cardiol 1993;21:1026–1027. 12. Levine RA, Jimoh A, Cape EG, et al. Pressure recovery distal to a stenosis: Potential cause of a gradient “overestimation” by Doppler echocardiography. J Am Coll Cardiol 1989;13:706–715. 13. Laskey WK and Kussmaul WG. Pressure recovery in aortic valve stenosis. Circulation 1994;89:116–121. 14. Schobel WA, Voelker W, Haase KK, Karsch KR. Extent, determinants and clinical importance of pressure recovery in patients with aortic valve stenosis. Eur Heart J 1999;20:1355–1363. 15. Baumgarter H, Stefennelli T, Niederberger J, et al. “Overestimation” of catheter gradients by doppler ultrasound in patients with aortic stenosis: A predictable manifestation of pressure recovery. J Am Coll Cardiol 1999;33:655–661. 16. Omran H, Schmidt H, Hackenbroch M, et al. Silent and apparent cerebral embolism after retrograde catheterisation of the aortic valve in valvular stenosis: A prospective, randomised study. Lancet 2003;361:1241–1246.