Unusual Mechanism of Coronary Perforation
During PTCA

Case Description. A 79-year-old female with history of hypertension, hyperlipidemia and peripheral vascular disease underwent successful direct stenting of her left anterior descending coronary artery on March 15, 2000 in the setting of unstable angina, utilizing a 3.0 x 15 mm Medtronic/AVE S670 stent deployed at 16 atmospheres (Medtronic-AVE, Santa Clara, California). Follow-up coronary angiography on September 1, 2000 for recurrent angina revealed angiographically significant in-stent focal restenosis (Figure 1). Intravascular ultrasound (IVUS) was performed to address the possible under-sizing of the stent, utilizing a 3.2 Ultra-cross catheter (Boston Scientific/Scimed, Inc., Maple Grove, Minnesota) By IVUS, the media to media vessel diameter was 3.6 mm proximal to the stent and 3.1 mm distal to the stent. Patient Management. Considering the IVUS measurements, a noncompliant 3.5 x 15 mm balloon catheter (Quantum Ranger, Boston Scientific/Scimed, Inc.) was selected to dilate the lesion. Markers were carefully positioned at the stent edges and the balloon was inflated. At a pressure of 4–6 atmospheres, the balloon would slide forward or backward, presumably due to the lubricity of the balloon coating and the focal noncompliant restenotic tissue. The balloon was then removed, inflated outside the body, and “scrubbed” with a dry gauze to remove any lubricious coating and improve its grip on the restenotic plaque. However, at an inflation pressure of about 6 atmospheres, the balloon continued to slide away from the target. A 3.0 x 20 mm Cross-sail balloon catheter (Guidant Corporation, Temecula, California), which was deliberately undersized, held its position on the target lesion when inflated to 14 atmospheres. However, residual stenosis was noted on the ensuing angiogram (Figure 2). The 3.5 x 15 mm balloon was then reinserted, positioned within the stent and inflated slowly. At about 12 atmospheres, the balloon rapidly moved forward 1 or 2 cm, coming to rest in a section of the left anterior descending artery that measured about 3.1 mm by IVUS. A Class II perforation was noted on the subsequent angiogram that showed pericardial or myocardial blushing (Figure 3). Heparin was reversed with protamine and the 3.5 x 15 mm balloon was left inflated at low pressure. A 3.5 x 20 mm Surpass balloon (Boston Scientific/Scimed, Inc.) was then used to perform an 8 minute inflation at 6 atmospheres, which appeared to seal the perforation. There was adequate perfusion through the Surpass balloon, evidenced by the absence of chest pain or electrocardiographic changes. The final angiogram confirmed that the perforation had been sealed and showed that a good angiographic result had been accomplished at the target site. A small outpouching had developed at the site of the coronary perforation (Figure 4). There was no evidence of pericardial effusion by echocardiogram or evidence of tamponade by right heart catheterization. The patient was discharged home in 36 hours without further complications. How Would You Manage This Case? Tim A. Fischell, MD Heart Center at Borgess Kalamazoo, Michigan In this case report, the operators have chosen a reasonable and relatively standard approach to the treatment of a focal in-stent restenosis. Historically, these focal in-stent restenotic lesions have a relatively good long-term prognosis when treated with plain old balloon angioplasty (POBA). Even without adjunctive intracoronary brachytherapy, the predicted recurrent restenosis rate for this lesion with POBA is in the 20% range. The long-term outcome could, perhaps, be improved somewhat with brachytherapy, with an expected recurrence rate of ~10%, but with some added costs and a potential for late complications. In balance, I believe that brachytherapy in this type of case is an option, but not mandatory. The assessment of the in-stent restenotic lesion with intravascular ultrasound is often enlightening. Stent under-sizing is the most common cause of in-stent restenosis. This under-sizing is typically only appreciated by IVUS. Thus, the operators choice of IVUS as a precursor to POBA is quite appropriate. Choosing the appropriate balloon size based upon the internal elastic lumina (IEL) to IEL IVUS measurements is a bit tricky. In a severely diseased segment with some compensatory remodeling, it may not be appropriate to choose a balloon size to match the largest media to media measurement. When stents are sized in this manner, there may be an unacceptably high incidence (1–2%) of coronary rupture. We would recommend a balloon size that is approximately half-way between the media to media IVUS measurement (in this case, 3.6 mm) and the angiographic (or IVUS) reference luminal diameter (in this case, 3.0–3.1 mm). Thus, an appropriate balloon size would have been a 3.25 mm diameter balloon in this particular case. Although this may seem like splitting hairs, using this approach in our lab has yield a 0% rupture rate in > 300 in-stent cases in the last three years. The other key choices in the treatment of this lesion relate to the frequent problem of “watermelon seeding”. All operators who have treated this type of in-stent lesion have experienced this vexing problem with today’s “lubricity coated” balloons. This was likely the main cause of the complication in this case, as the balloon squirted forward into the smaller distal left anterior descending coronary artery segment. This balloon movement during high-pressure inflation is particularly troubling during the staging of intracoronary brachytherapy, since it can extend the injured length and contribute to “geographic miss”. For all of these reasons, we have shifted to the nearly exclusive use of the cutting balloon for these in-stent lesions. Alternatively, one can create a poor man’s cutting balloon by placing a second, “buddy” wire alongside the conventional balloon, which will often eliminate this balloon slipping problem. Debulking with rotational atherectomy, or perhaps laser, may also reduce the risk of balloon “watermelon seeding”. Finally, let’s comment briefly on the management of the coronary perforation that ensued. The operators were fortunate that no glycoprotein IIb/IIIa inhibitor was used. In our experience, even minor, previously benign guidewire perforations can be very persistent with these antiplatelet drugs on board. The key ingredients to the successful management of coronary perforation were followed in this case. First, immediately inflate a balloon at low pressure to prevent hemo-pericardium and tamponade. Next, reverse the anticoagulation. Perform a very prolonged balloon inflation (perfusion balloon is ideal if available) at low pressure to attempt to seal the perforation. In this case, the maneuver was successful. If the tear cannot be sealed, as is often the case with a true balloon-induced rupture, the next approach should be to place a covered stent (Jomed) at the site. This bare-mounted device should be mounted on a non-compliant balloon (preferably one without a lubricity coating in order to minimize stent embolization). Finally, if delivery of the covered stent fails, it will usually be necessary to consider surgical repair versus prolonged intrapericardial drainage with a pigtail catheter. With the proper techniques, the incidence of coronary rupture and the resultant morbidity and mortality can be minimized. Ted M. Parris, MD Hahnemann University Cardiac Catheterization Office Philadelphia, Pennsylvania In the above case report, the authors describe an unusual mechanism of PTCA-related coronary perforation. The case presumably reflects a focal perforation from “watermelon seeding” of an oversized balloon distally into a segment of vessel whose reference diameter was smaller than that of the inflated balloon. Given the relatively minimal size mismatch, it’s surprising to me that the result was a perforation rather than a dissection. Once the perforation was recognized, was seen to be Class II,1 and the patient did not develop acute tamponade or hemodynamic compromise-judicious conservative management with a prolonged balloon inflation “sealed” the perforation. This is the appropriate and preferred treatment (i.e., prolonged balloon inflation ± stenting) and effective for most Class II perforations. My management, once the perforation occurred, would have been identical. Could this unusual complication have been avoided? Granting the fact that hindsight is 20/20 (particularly for those not involved with the case), let’s see if another approach to this restenotic lesion might have been considered at certain points in this case. Our general approach to focal in-stent restenosis has been repeat balloon dilatation (now often with adjuctive brachytherapy), and is quite safe and effective. Thus, my initial approach would also have been repeat PTCA. However, this lesion was a bit unusual. The authors suggest that the “watermelon seeding” effect was from the lubricity of the balloon coating and the focal non-compliant restenotic tissue. The fact that “scrubbing” the lubricious coating did not alter the situation suggests to me that the more critical factor may have been lesion- rather than equipment-related. I think we’ve all encountered these types of lesions sometime during our interventional practice. What might be some alternative approaches? Possibilities include the following: • Use of an even less compliant balloon catheter with thicker, more poorly rewrapping balloon material (e.g., SciMed NC Ranger). This balloon reaches its nominal size quickly at lower pressures than the balloon initially employed and may have had less of a chance of “squirting” distally. Given the case as presented, I might have considered this balloon material once recurrent failure occurred with the Quantum Ranger. • Some of these “non-compliant” lesions within stents, even focal, may be better treated initially with a “debulking” device such as a cutting balloon. While initially employing simple PTCA, I would probably have switched to a cutting balloon (3.25 or 3.50 mm x 10 mm) or Rotablator burr at 2 decision points in the case: 1) when the first balloon still “squirted” after removing the lubricious coating, or 2) when a residual lesion was still apparent after inflation with the 3.0 x 15 mm Cross-Sail at 14 atm (effective balloon diameter = 3.29 mm). As stated, once the perforation occurred, it was quickly recognized and expeditiously managed with virtually no subsequent morbidity. I commend the authors for their fine work on a potential “disaster” of a case. References 1. Ellis SG, Ajluni S, Arnold AZ, et al. Increased coronary perforation in the new device era. Incidence classification, management and outcome. Circulation 1994;90:2725–2730. Bernhard Meier, MD University Hospital Swiss Cardiovascular Center Bern Bern, Switzerland The depicted complication of a coronary rupture during balloon angioplasty for an in-stent-restenosis is the product of a questionable policy and a well known phenomenon. The generous balloon sizing based on intravascular ultrasound measurements is safe if used inside the stent. As soon as the balloon protrudes outside the stent, particularly in the distal, smaller portion of the vessel, the ultrasound-driven balloon sizing becomes dangerous as illustrated. Balloon selection by ultrasound measurements is fairly safe as long as the generally large balloon is not used at high pressure where balloon compliance makes it even larger. Redilating in-stent restenosis almost invariably requires high pressure. In addition, it is not uncommon that it is difficult to stabilize the balloon within the stent during inflation. The possibility that the balloon is squeezed distal to the stent while inflated at a high pressure (melon seed phenomenon) is a real one and materialized in this case. Dissection and even rupture, though rare, may occur. Working with small caliber guiding catheters (5 or 6 French), it is usually possible to advance the guiding catheter fairly deep into the coronary artery, close to the stented segment, to be redilated. During inflation, the guiding and balloon catheters can be controlled bimanually by the operator;therefore, slippage can be avoided. Once the rupture has occurred, it is important to distinguish whether it is a free rupture emptying into the pericardium or a covered rupture. The latter can be recognized by a blush of dye around the dilated segment that does not clear rapidly. Fortunately, this was the case in the patient reported. A free rupture mandates the use of a covered stent. If the covered stent saddles the ruptured segment, the hole is sealed and flow is maintained. If a covered stent cannot be brought up to the rupture (e.g., small peripheral vessel), a covered stent might be placed into the main vessel at the take-off of the ruptured vessel to seclude it from inflow. This does not work if the ruptured vessel is copiously collateralized. In that case, or to avoid the use of a covered stent in a healthy vessel, the ruptured small vessel can be clogged by coil embolization. In the presented case, the use of a perfusion balloon for 8 minutes was a good choice, but the case would probably have turned out well even without any further measure. Thomas J. Linnemeier, MD, FACC Guidant Corporation Menlo Park, California This 79-year-old female developed in-stent restenosis of her left anterior descending artery (LAD) five and one half months after the original stent procedure, where a Medtronic/AVE S670 stent (Medtronic/AVE, Santa Rosa, California) was implanted. Based on one angiographic view, the LAD appears to be diffusely, though concentrically, restenosed throughout most of the length of the stent. If treated with repeat angioplasty alone, the odds of in-stent restenosis a second time is probably in the order of 50%, with successful radiation therapy, probably less than 10–15%. I would pre-treat the patient with ASA, clopidogrel, heparin (monitoring the ACT in the 200 range), and one of the GP IIb/IIIa platelet inhibitors, depending on her angina status and if she were diabetic. I’d then place an 0.014´´ wire to the apex of the LAD and pre-dilate with a 3.0 mm cutting balloon (Boston Scientific/SciMed, Maple Grove, Minnesota) in order to avoid watermelon-seeding. I’d then apply radiation therapy from one of the three commercially available radiation therapy companies, the Check-Mate system (gamma radiation, Johnson and Johnson, Warren, New Jersey), the Novoste Beta-Cath system (beta radiation, Norcross, Georgia), or the Galileo system (beta radiation, Guidant, Santa Clara, California), being absolutely sure that both the proximal and distal areas that were injured by the cutting balloon, were also radiated. If on final angiography there was a