The Surgeons

Chapter 8 THE FUTURE OF HEART SURGERY

THE ANNUAL TCT, OR TRANSCATHETER CARDIO-VASCULAR Therapeutics, conferences must be the biggest, slickest, most spectacular medical gatherings in the world. The one I attended, the weeklong TCT2006, held at the cavernous Washington (D.C.) Convention Center in late October 2006 could boast of 11,000 attendees, most of them cardiologists, from eighty countries. The main opening event, held in three large ballrooms with retractable walls removed, had seating for 3,000, but was still lined with standees. Two other large opening events plus multiple smaller ones were in session at the same time. The multimedia wall display in the main room was perhaps a hundred feet wide, with a curved black-leather-and-rosewood conference table stretching thirty or forty feet along its base.

The founder and guiding spirit of the TCT is Martin Leon, who was welcoming his guests from the ballroom podium. He is fifty-five, personable and low-keyed, and one of the most influential interventional cardiologists in the world. Interventional cardiologists thread catheters, or long, thin tubes and wires, through blood vessels to get access to the heart. Using tiny, catheter-based tools and devices, they can perform many tasks, such as reopening an occluded coronary artery, that were once the exclusive domain of open-heart surgeons. Leon, an outstanding practitioner and innovator, is a former director of interventional cardiology at the NIH and founder and chairman of the Cardiovascular Research Foundation. His day job is as associate director of the interventional cardiology program at Columbia-Presbyterian.

After Leon’s opening remarks, the wall display dissolved to the catheter lab at Columbia-Presbyterian, where a team headed by his partner and Columbia interventional cardiology director, Jeffrey Moses, was about to perform a difficult procedure on a real patient, using a “rotablator,” a miniaturized tunnel-digging machine. As the operation proceeded, multiple windows on the wall display zoomed seamlessly from teamwide shots, to meticulous closeups of the catheter insertion, to images of the catheter threading its way through the arterial system. Windows popped open and closed as shoptalk bounced back and forth between Moses’ team and Leon and the twelve other famous cardiologists at the main table—about rotablator rpms, balloon pressures, wire diameters. At one point, when the rotablator got stuck, Leon looked almost happy. “Okay [addressing the audience], this is a great teaching case. You have to realize these are difficult procedures. Now, Jeff, how will you get that out?”*

The rotablator was freed, but the case was far from finished when the Columbia team waved good-bye, and the display wall dissolved to a transatlantic cartoon trajectory to Milan, swirling shots of La Scala and other landmarks, “Grand March” trumpets from Aida, then Antonio Colombo and his team at San Raffaele Hospital waving “Hello, Martee.” Leon and the other head-table cardiologists kibitzed enthusiastically as Colombo placed a tricky V-shaped stent, or wire-mesh support, in a diseased arterial branch point. But precisely at thirty minutes, the room was teleported away, with a flourish of samba and clips of Carnaval revels, to an operation in São Paulo, and from there, finally, to one in Rotterdam. All the procedures were cutting-edge, all the teams were world-class, and all of them were clearly friends of Leon.

The several days I was there followed that pattern—dozens of events, packed rooms, flawless tech-dazzle. Sponsor banners from drug and device companies festooned the halls, and one large floor was mostly given over to company displays. All of cardiology’s industrial giants were there—GE, Siemens, Philips, Medtronic, Toshiba, Cordis, Boston Scientific, Abbot Vascular, Edwards—plus an army of hopefuls, like Bioheart (cell-based heart regeneration), Atritech (a clot catcher), Cardia (tool to close septal holes), and many others. The pharaonic assemblies, the parade of companies, the sheen of serious money, all drove home that interventional cardiology is “hot.”

And that cardiac surgery is very much on the defensive.

I first met Leon when I sat in on a meeting between him and Craig Smith in early 2006. The meeting was ostensibly an administrative one—finding space to create a “hybrid OR,” equipped for both cardiac surgery and interventional cardiology. But it was really about the future of heart surgery.

Interventional cardiology developed more or less side by side with the boom in cardiac surgery starting in the mid-1960s, initially as a diagnostic handmaiden, but as the TCT demonstrated, increasingly as a competitor. The first cardiac catheters were used to measure blood dynamics within the heart, then to pinpoint blockages by squirting dyes through suspect arteries. But cardiologists discovered that they could use the catheter to push aside coronary occlusions, and catheterization became a first-line defense after a heart attack. The next steps were to expand constricted arteries with tiny balloons, then more permanently with metal stents. Catheter-based tools are now progressing very rapidly. At least one company is marketing a micro cardiac pump that can drive a full complement of blood, in effect, a button-sized proto artificial heart on a wire.

The discussion between Smith and Leon was an easy one—Smith is too rational, and Leon too emollient, to imagine it otherwise. They pretty quickly settled on a boundary line between cath lab procedures and those that required a full-blown cardiac surgery OR and staffing. They also agreed, resources permitting, to create two hybrid ORs that would support either cardiac surgery or catheter interventions, and reached a rough consensus on capital requirements. The cardiologists need walls full of catheters and guide wires, and far more elaborate imaging equipment than surgeons use, while the surgeons’ open-chest working environments impose the most demanding sterility requirements and a staff quotient tuned to the probability of disaster. (Cardiac cath labs ship their disasters to a surgical OR.)

The hybrid OR was completed just about the time of the TCT conference. The hospital was willing to support only one, not two, and they chose to locate it within the cardiology suite, a not-so-subtle signal of shifting specialty pecking orders. The hybrid OR is ideal for the more invasive cardiological procedures that require surgical-quality sterility and backup, as well as for true hybrid operations—I’ll describe an important one later—that incorporate both surgical and catheter-based techniques. The longer-range objective is that a hybrid procedural environment will speed the evolution of a new breed of “cardiac interventionalist” trained in both surgical and interventionalist disciplines. The Columbia surgeons and cardiologists have already established one of the first combined residencies in interventional cardiology and cardiac surgery. Mat Williams, the first joint fellow, who completed his cardiothoracic surgical residency in 2005 and his interventional cardiology training in 2006, now holds a joint appointment as a Columbia-Presbyterian interventional cardiologist and cardiothoracic surgeon, with full privileges in both departments, the first such position in the nation.

For a cardiac surgeon, Smith has been unusually open-minded toward the advances of the cardiologists, and seems to have become one of their favorite surgeons—he was one of the few surgeons on a panel at the TCT conference, for instance. His more closed-minded brethren, however, might protest that he has the luxury of an overflowing patient registry, while many of them are struggling just to stay afloat.

The Plight of the Cardiac Surgeon

Consider the CABG, for the last forty years the bread-and-butter operation for the average heart surgeon. During the decade 1993–2003, the number of CABGs in the United States declined about 11 percent, from about 340,000 to just over 300,000, even though the prime-age patient population grew by more than a fourth. Interventional cardiology, in the meanwhile, has been struggling with exploding workloads. Stents and angioplasties, the main catheter-based CABG substitutes, had more than doubled, to more than a million, by 2006. By 2000 or so, one-and two-coronary bypass operations were already becoming a rarity.

Rubbing it in, insurance reimbursement rates for CABGs have been steadily falling as well. Medicare reimbursement for the standard leg-vein triple-bypass operation is down 20 percent since 1998, or by about a third if inflation is taken into account. Private insurers pay better rates than Medicare, but the gap is narrowing fast as health plans consolidate and demand hospitalwide price concessions. Fees for interventional cardiologists have been dropping even faster than for heart surgeons, but soaring volumes more than make up the difference. Certainly, the TCT conference gave no hints of an embattled profession.

Postings on CTSNet, a Web site for cardiothoracic surgeons, are almost elegaic. Here’s one from a surgeon in private practice in Missouri:

No jobs in Cardiac Surgery[?] Is this just a fallacy or is there any truth to it? Well, just look around and talk to your colleagues…. People are looking for jobs which are just not there at all, regardless of your training, years of research, graduating institution…. Obviously either the training was not adequate or the requirements of the market have changed. After an average of 10 years of cardiothoracic training, surgeons are opting to go for a second specialty training in anesthesia, vascular surgery and plastic surgery. The starting salary for [cardiac] surgeons is somewhere between $120,000 and $240,000, whereas an anesthesia graduate with a pain fellowship starts in the $400,000 range. [In some cases,] you are better off working at one of the larger institutions as a superfellow/adjunct-surgeon/instructor or even work as a physician assistant!…. A recently advertised job in South Florida offered the handsome compensation of 80,000/year for fully trained cardiac surgeons.

Signs that the wolf is really at the door came in the spring of 2006, when seventeen newly minted cardiothoracic surgeons failed to find jobs—the first time that has ever happened. The accreditation authorities responded by cutting back on accredited programs and residency slots. Worse, there is evidence that the best candidates may be drifting toward other specialties. Cardiothoracic residency slots have always drawn a surplus of candidates. But in the 2006 national residency match, there were only 102 applications for 126 approved slots. Eleven of the applicants then dropped out before the match, so 35 slots were not filled. That is unheard of—like Harvard going begging for undergraduate applications. The effect on academic program staffing will be profound. In the academic year 2004–2005, the first-year residency class was 141. In 2007–2008, when the group from the 2006 match begin their residencies, there will be only 91, a drop of 35 percent.

The relentless march of the cardiologists has not been without setbacks of its own. The first coronary “angioplasty” (or “vessel remolding”) in a human took place in 1977. A catheter tube and balloon is threaded to the point of the occlusion, and the balloon is inflated to push the arterial walls open. It takes only an hour or so, usually with only local anesthesia. Patients go home the next day and are back to work a few days later—no chest scars, no months of recovery, no worry about bypass-machine neural effects. Angioplasty spread like wildfire. But as the case registry grew, the data showed a disturbing tendency toward reocclusion. So, yes, angioplasty is a patient-friendly way to reopen a coronary artery, but the odds are high that within a few years you’ll need a do-over. With the CABG, by contrast, although the initial insult is much worse, the repair is closer to permanent. For most patients, the life-extension effects of the two procedures are roughly the same. Over time, cases with standard one and two-vessel disease have migrated steadily to the cath lab—two or three redos over a decade or so isn’t a bad tradeoff against a CABG. Three-vessel disease has mostly stayed in the CABG camp—three angioplasties imply a lot of redos—and there are subsets of patients, like diabetics and people with occlusions around the left ventricle, who clearly do better with the CABG. Cost differences are actually fairly modest—the initial angioplasty is much cheaper than a CABG, but the necessity for further interventions pushes total costs toward convergence.

The wire-mesh stent, unveiled in 1994, looked like a solution to the reocclusion problem. It is delivered folded around an angioplasty balloon and expanded within the occlusion to create a permanent scaffolding. But the stent’s advantage over traditional angioplasties proved to be disappointingly modest. Yes, the scaffolding prevented the arterial walls from recollapsing, but the presence of the stent induces scar tissue proliferation, which is itself an important source of occlusions.

The drug-eluting stent—a stent coated with medication to prevent scar tissue formation—was the next break-though, because it sharply reduced the rate of reocclusion. The DES, as it’s called, burst on the scene in 2003, and by 2006 accounted for more than 80 percent of all U.S. stent placements. In just the three years since the FDA’s approval of the first DES, a Johnson & Johnson product, the market exploded to annual sales of $6 billion, amid a colossal market slugfest between the two dominant players, J&J and Boston Scientific. With sales figures that rival those of blockbuster consumer drugs, DES is now the richest medical device market in history. I went to a conference in the spring where speakers confidently forecasted the near-term demise of the three-vessel CABG.

A few months before the 2006 TCT conference, however, new data associated the DES with small, but statistically significant, increases in longer-term mortality. Excessive blood clotting in later years, it appears, may fully offset the benefits from low early reocclusion rates. One prominent cardiologist labeled the widespread use of DES “an epidemic of madness.” By the spring of 2007, DES usage in the United States dropped from nearly 90 percent of all stents, by far the highest in the world, to about 70 percent, or about the same as in Japan. European stent usage, which was already low at 54 percent, dropped slightly to 51 percent, although there was wide variation from country to country. In December 2006, a divided FDA advisory committee concluded that a DES was superior to a bare-metal stent for the relatively straightforward conditions approved by the FDA. For the more complex off-label uses—patients with long lesions, multivessel stenting, bifurcation (branching) lesions, patients suffering from acute myocardial infarctions (heart attacks), and diabetics—that comprise the majority of DES placements, they agreed only that little was known of the efficacy and risks of DES compared to bare-metal stents or surgery. The panel also strongly recommended an extended anticlotting drug regimen for off-label patients and discouraged DES in patients who might have difficulty maintaining such a regimen.

The lavishness with which scientific brainpower and financial capital is being showered on the stent, however, as well as the large number of commercial teams working on the challenge, virtually ensure that stents will continue to expand their franchise. The cardiac surgeon who assumes his triple-bypass franchise is safe is reading the wrong tea leaves.

In the Cath Lab

After I had spent several months with the cardiac surgeons, I arranged with Marty Leon and Mat Williams to learn more about cath labs. At the time, Williams was assisting one of the senior cardiologists, Michael Collins, with diagnostic catheterizations. The “cath lab” looks like a mini-OR, but without the sterility trappings—for most procedures, everyone is in the usual OR scrubs, but there is no sterile field, no gowning, none of the obsessive elbow-deep prescrubbing. Inserting the typical cath looks much like giving blood at a Red Cross station. But while blood collection needles travel only a few inches into a vein, diagnostic catheters are threaded all the way into the heart, or liver, or other organ.

Every cath lab has an enormous assortment of guide wires and catheters arrayed along its walls. A routine angiogram usually requires the insertion of just a single catheter and guide wire, but in more complex procedures the guide wire may be left in place to facilitate access for a variety of other catheters and tools. Size, stiffness, outer coatings, may all vary depending on the instrument or the patient. There is considerable skill involved in advancing a catheter—how hard can I push in this patient to get around a difficult corner? Soft wires help in traversing sinuous regions, but stiffer wires may be needed to push through an occlusion. The shape of the catheter tips, the nature of the coatings, weight and stiffness, are the subject of intensive research. One firm advertises its products’ superior “pushability.”

The first diagnostic cath I saw was a straightforward angiogram. The patient was being considered for a lung transplant, and Collins was just confirming that he had no coronary disease. The catheter was inserted into the left femoral (leg) artery in the groin area. From there it is a fairly straight shot through the right iliac artery, the abdominal aorta, the thoracic aorta—all just locational labels on the same giant pipeline—up to the aortic arch, then through the arch down to the aortic root and the junctures of the right and left coronary arteries. Soft tissues don’t register on the imaging screen, so as Collins advanced the catheter, it appeared to float in a strange curving pattern through an empty rib cage, moving and bending of its own volition. While the image helped Collins fix his location, as an experienced interventionalist he could have done it by feel. At the aortic root, Collins rotated the catheter tip until he felt the entrance to the left coronary, inserted the catheter, squirted a dye, and froze the image as the dye flooded through the left-heart arterial network. He withdrew the catheter out of the left coronary, rotated it again, found the right coronary, and repeated the process. The coronaries looked fine.

The second procedure used a Swanz-Ganz catheter, a device that’s been around since 1950. I’d seen the Swanz-Ganz before, since anesthesiologists use it to monitor cardiac flow dynamics in almost all open-heart surgeries. Collins’s patient was an older woman with multiple comorbidities and progressive heart failure. The Swanz-Ganz has two blood-pressure sensors, one at the end and another further back, as well as a dilating balloon. It is usually inserted through the jugular, then follows the venous blood flow through the heart toward the lungs—from the jugular to the vena cava, and through the right atrium and right ventricle into the pulmonary artery. In its final position, the intermediate sensor is in the right atrium, and the end-point sensor is in the pulmonary artery. In this case, Collins was primarily interested in the end-point measure. He inflated the balloon at the endpoint sensor, briefly blocking blood flow through the pulmonary artery into the lungs, so the only pressure registering at the sensor was the backwash from the lungs’ blood pressures. That was how the cardiologists in Iowa and St. Louis first identified Erika Maynard’s pulmonary hypertension.*

The reading, as Collins had feared, showed pulmonary hypertension. He kept the catheter in place while he had the patient breathe pure oxygen, then, after waiting a bit, introduced a nitrous oxide (laughing gas) mixture and took another reading. If the lungs were at all reactive, the catheter would have registered a noticeable drop in pulmonary pressures. There was almost none at all. Collins looked sad. This was a very nice woman, he said, and he was going to have to tell her that she was going to die and that there was not much they could do for her. Her age, her physical condition, and the hypertension in the lungs pretty much ruled her out as a transplant candidate.

Later, I caught up with Leon to watch a stent placement. The patient had three blockages, but none in the major left-ventricle arteries, so he had opted for stents rather than a CABG. Leon started by threading two catheters into the right coronary. One was for angiogram dye, while the second placed the guide wire that would be used to transport the stents. He worked from two large imaging screens. When he squirted a bolus of dye, the right screen froze a shot of the bramble-bush layout of the patient’s right coronaries, pinpointing the occlusions. The left screen was to track the stent catheter. As Leon worked the guide wire toward the first occlusion, it struck me what a remarkable feat of visual-tactile intelligence I was watching. Although everyone’s coronary arteries follow roughly the same layout, the detailed branchings are unique. As he pushed the stent catheter, Leon could glance at the angiogram shot on the one screen and relate it to the wiggle pattern of the disembodied guide wire on the other, picking out turns mostly by the feel of the wire. Every couple of minutes, he’d pause for a pop of dye on the left-screen image to confirm that the wire was where he thought it was.

The first occlusion was a long one, located in a twisty byway, with a nearly right-angle turn in the center of the diseased section. For almost forty-five minutes, Leon pushed up, and then withdrew, several different stenting solutions. Drug-eluting stents are stiff, and he couldn’t negotiate the turn with one of the proper length. He finally succeeded in placing two separate stents in the two legs of the right angle, eliciting quiet cheers from the whole team. The last two stent placements took only about five minutes, and the catheters were withdrawn, the wound bandaged, and the patient ready for transport within another quarter hour or so. Although the patient was sedated and drowsy, he was awake and communicative throughout the procedure.

The Cribier-Edwards Valve

Stan Rabinovich, voluble, rosy-cheeked, and grandfatherly, struck me as a happy man when I met him at his one-person office in Englewood, New Jersey. Rabinovich, along with Marty Leon and several other partners, was a founder of a company called PVT, in 1999. A couple of years before my visit, PVT had been purchased by the California-based Edwards Lifesciences, an important manufacturer of cardiology supplies, netting $6 million for each of the founding partners, along with the prospect of additional payments down the road. An extra fillip for Rabinovich is that Edwards has retained him as a one-man evangelist for the product he helped create, and which he loves to talk about—a catheter-delivered aortic valve replacement.

Rabinovich is an electrical engineer. Along with a friend, Stan Rowe, from Johnson & Johnson, he made major contributions to stent manufacturing. Rabinovich was sent to Paris in 1995 as head of J&J’s European business development, where he met Alain Cribier, a well-known French cardiologist, who had pioneered the technique of “valvuloplasty”—the use of an angioplasty balloon to loosen and expand a calcifying aortic valve. Valvuloplasty was a disappointment, since so many valves quickly recalcified. Cribier had moved on to a valve replacement delivered “percutaneously” (doctor-Latin for “through the skin”—i.e., by means of a catheter). Rabinovich brokered a development agreement between Cribier and J&J.

Rabinovich and Rowe had both left J&J by 1999, when Rabinovich got a call from Cribier. Cribier told him that J&J “wasn’t doing anything with his valve,” and asked for his advice. Rabinovich and Rowe had met Leon when he was an outside medical director for J&J’s interventional business, and took a train to Washington, where Leon practiced, to ask whether he thought Cribier’s invention was a good idea. Leon thought it was a very good idea, and the three quickly formed a partnership with Cribier, who reacquired his rights from J&J. They put together a business plan and raised $500,000 from early-round “angel” investors, mostly through Leon’s contacts in the medical community. Rabinovich and Rowe left their jobs and rented the one-room office where I met Stan. Impressively, they never hired other staff—no receptionist, no bookkeeper, no research assistants. The name PVT stood for Percutaneous Valve Technology.

They discovered they had a patent problem. A Danish doctor had patented a solution much like Cribier’s in 1985, and his rights had been purchased by an American heart surgery company, Heartport, which had built a strong patent position. “We spent a long time analyzing the Heartport patents,” Rabinovich said, “and finally decided there was no way around them.” Heartport marketed tools and devices for performing heart surgery through small incisions. But they thought like surgeons, Rabinovich said, and gave up on the valve patent because they couldn’t figure out how to remove the old valve. Cribier, by contrast, with his experience in valvuloplasty, left the old valve intact; he just expanded it with a balloon and put the new valve inside it. The pressure of the old valve ring would keep the new one in place while normal scar tissue created a permanent bond. Over a few hectic months, they struck a reasonable deal for the Heartport patent, located a high-tech contract manufacturer in Israel with good capabilities in materials sciences, and began developing valve prototypes. The technical challenges were formidable—the stent containing the valve in the current model is reputedly the strongest stent ever made.

PVT’s world changed in 2002 when Cribier executed the first successful catheter-delivered aortic valve replacement in a human, working under a French “compassionate care” exception on a terminally ill patient who could never have survived standard surgery. Medical press front pages all around the world featured a picture of Cribier with his patient, who was smiling broadly and sitting up in bed, chest intact, drinking champagne within hours of receiving his new valve. Soon after, PVT executed a venture capital financing for $5 million, followed by a $14 million round that included industry heavy-hitters like Medtronic. Buyout offers started in earnest in 2003, and a deal was done when Edwards showed up with a can’ t-refuse offer late in the year. Rowe is now running Edwards’s valve business, while Rabinovich holds forth contentedly in Englewood. Not surprisingly, there are now dozens of companies working on a variety of aortic valve solutions, and nearly as much activity in mitrals, including some advanced work at Columbia.

An invasive heart device is subject to the same kind of rigorous approval processes as a new drug. At the time of the TCT conference, early-stage Cribier-Edwards trials throughout Europe and the United States had performed 282 procedures. Trial data were available only on the 55 cases conducted at Columbia-Presbyterian. Valves were successfully placed in 48, or 87 percent, of the patients, and the thirty-day mortality rate was 7.3 percent. That was more than double the most recent statewide mortality rate in surgical aortic valve replacement, but still a superb outcome, since the trial was limited to patients who were too sick or frail to tolerate standard surgery. Leon successfully placed a Cribier-Edwards valve in a hundred-year-old, as well as in several patients in their nineties.

I saw a Cribier-Edwards placement in April. Enthusiasm for the valve was building, so the operating area was unusually crowded—there was at least one team of observers from Europe, several senior figures from the hospital, people from Edwards, and probably others. The patient was a male, in his seventies, showing a reops’ steel chest sutures on the imaging screen, who had been turned down for a surgical valve replacement. Leon and Moses were the interventionalists, with Mat Williams in attendance as well. The procedure took place in a cath lab, but one that was outfitted with most of the trappings of an OR—there was sterile field, the team was in sterile gowning, and the attending anesthesiologist, Sanford Littwin, was from the cardiac surgery anesthesiology team. I stayed in the windowed control room, which offered a good viewing perspective, with a complete bank of imaging screens, a microphone feed from the table, and a nurse–cardiology technician who was recording every step of the operation. It also happened to be one of the few operations that were unsuccessful, although it is a dramatic illustration of the procedural challenges.

To install the valve, the interventionalist first threads a balloon catheter from the femoral artery up through the aortal arch and into the valve. The balloon is pushed into the aortic valve and dilated—a valvuloplasty—to clear away calcification and expand the valve opening. In the case I saw, they did it twice. The balloon catheter is then withdrawn and replaced by the valve catheter. A challenging feature of the Cribier-Edwards procedure is that it requires an unusually large catheter payload. The patient I saw had a fairly large 26 mm valve opening. The Cribier-Edwards 26 mm valve-stent can be “crimped” down to about a third of that diameter, or just small enough to get through most adult male femorals. Once the new valve reaches the aorta, traverses the arch, and is positioned just above the old aortic valve, the patient’s heart rate is electrically accelerated to about 200 beats per minute, inducing a fibrillation. Instead of beating, the heart just buzzes in place, so there is a stable catheter target. The new valve is quickly centered within the old one and expanded with a balloon to lock it into position. Then the balloon and catheter are withdrawn, and the pacing is decelerated back to a normal rate. The fibrillation lasts for only a half minute or so, which is reasonably safe.

In the case I saw, everything went as planned until the insertion of the valve catheter. This was a patient with extensive calcification in both femoral arteries, although one of them seemed clear enough for access. The valve catheter traversed the femoral smoothly, but got stuck on an outcropping of calcium in the iliac artery, around the mid-trunk area. Jeff Moses was inserting the catheter, and for several minutes he twisted or turned, angled and pushed it, but he couldn’t clear the calcium. Then he withdrew the valve and dilated the calcified spot with a 10 mm balloon. (It’s larger than cardiac catheter balloons and is normally used in noncardiac vascular procedures.) When he reinserted the valve cath, it hung up for a few seconds at the calcified spot but, with a little pushing, suddenly released and sailed through, amid some back-slapping in the control room.

Then someone from the table said, “Arrest!” A glance at the EKG showed the heart beating, but blood pressure at zero. An injection of angiographic dye told the story—there was an evil plume of blood on the image screen, pouring out from the ruptured iliac like smoke from a burning building, right from the location of the calcification hang-up. One of the hospital’s senior doctors, who had come down to the control room to greet the Europeans, put both hands on the table in front of him, closed his eyes and bowed his head.

Everyone in the room, I believe, thought they were witnessing a mortality. There had been several iliac ruptures in the European trials, and every one of the patients had died. This patient not only survived, but did so without apparent ill effects—a tribute to the skills of the people around the table. Somehow, Leon and Moses got another big balloon to the point of the rupture and blocked the bleeding. Littwin revved up the patient’s blood pressure but managed not to dislodge the balloon. Williams did a fast cutdown to the site of the rupture, clamped it, and executed a rapid repair. Within the twenty minutes or so it took to ready a cardiac OR, the patient was already stabilized. Williams later told me that he worried that ischemia damage might require a leg amputation, but the patient was up and walking by the second day.

To my surprise, and to Marty Leon’s infinite credit, I saw that same image again some months later at the TCT conference—the same dreadful shot of dark, dyed blood blowing out of a raggedly torn iliac—when he showed it to a room packed with about five hundred cardiologists. He had just completed a glowing account of the Cribier-Edwards trials—the high success rate, the stability of the valves, their excellent performance and tight sealing. Then he said, “Most people just like to show you what works. But I like to show everything that can go wrong.” And he flashed up screen images of all of his failed cases, occasionally even lingering over them. “This is an extremely promising procedure,” he concluded, “but it’s not yet ready for general practice.” There were a lot more trials to go, more work to be done on access methods, more development work on smaller, more “pushable” valve transporters. But there was no doubt that the craft of valve replacement was heading for radical change.

Meanwhile, Back at the OR

Not all the new developments are in the cath lab. In the spring, I arranged with Mike Argenziano to see a “robotic” surgery. The cardiologists had referred a patient with two diseased arteries—one they could stent, but the other, involving a large area of disease in a major left-side artery, required a bypass. Since the patient was hoping to avoid a sternotomy, it was agreed that the surgeons would do a minimally invasive off-pump bypass on the critical artery, while the cardiologists would stent the second artery a week or so later. Smith would execute the bypass, and Argenziano would harvest a mammary artery, using the da Vinci “robotic” surgical tool.

The da Vinci is a massive machine with two major components. The operating unit, which comes on heavy wheels and is more than six feet high, is positioned at the foot of the operating table. It has up to four large, praying mantis–type stainless-steel arms with a video display at the center. The control unit is about the size of a large armchair and is positioned off to the side of the OR. The surgeon sitting at the console can control any of the arms through two control knobs, while watching what he’s doing on the console’s 3-D, high-definition color screen. The table team can also follow the surgery through table video display.

The two critical features of the da Vinci are the 3-D picture—depth-perception mistakes in surgery can be disastrous—and the “endo-wrist” knob system. The knobs transmit the surgeon’s actual wrist and finger motions to the remote tools, in user selected transmission ratios. Argenziano sets his at 3:1—a 3 mm turn of his wrist turns the tool 1 mm, while squeezing index finger and thumb will open and close the blades of a scissors one-third that much. (Smith, characteristically, sets his at 1:1.)

I came into the OR when Argenziano was setting up the da Vinci. He was being helped by a woman—in her late thirties, I guessed—wearing green OR scrubs rather than Columbia-Presbyterian’s light blue. Her name was Felicia Brodzky, and she was the local da Vinci account manager. We met for coffee a few days later, and she explained that she spent her workday traveling from OR to OR throughout Manhattan, helping surgeons with the da Vinci. Her boss would prefer that she wear a business suit, but she said she saved a lot of time by wearing scrubs to work, partly to avoid all the changing, but also because security staff automatically wave through people in scrubs—a trick I had figured out by then too. She wasn’t a medical professional, but her undergraduate work was in science, and she was clearly very knowledgeable about surgery. In the two da Vinci procedures I saw, she was very useful, jumping in when a resident got confused during a tool change, for instance. Her company is doing well, and I imagine she is well paid.

Argenziano used three of the arms for the arterial harvest—one for a camera, and two for tools. The knob settings are adjustable, so either one can control any of the tools. The resident made three quarter-inch incisions around the patient’s chest and inserted a tool arm through each one. Then the patient’s chest was inflated with gas to increase the working space, and the operation proceeded as normal. Argenziano arguably had better visualization of his target vessel than he would in an open-chest operation, he could zoom his vision field in and out as he needed to, and he had extremely fine, tremor-free control over his instruments. The functionings of the da Vinci that were not directly related to the surgery—like getting the arms positioned at setup—struck me as a bit clunky, and the size and expense of the machine makes it a difficult purchase, but the surgery itself proceeded beautifully. There was a good selection of tools, and Argenziano could cut, clamp, grasp, and stitch virtually any way he pleased. When he finished, the mammary had been cleanly harvested and clamped, and was lying on the pericardium for Smith’s bypass.*

The minimally invasive bypass was routine. Argenziano cut a four-inch incision between two ribs, which he spread with a miniretractor, but without breaking them, and Smith used a long set of tools to execute an off-pump bypass. The bypass itself probably took a little longer than a single open-chest bypass would have done, but the overall savings, in chest opening and closing, trauma recovery, and bypass machine side effects, must have been immense. This was also the first time I’d seen Argenziano’s “A-rod” in action. It is a rodlike tool he invented to stabilize a coronary artery during a minimally invasive off-pump operation, and the name both reflects his authorship and proclaims his favorite Yankee player. A major surgical supplier carries the A-rod in its catalog.

In December 2006, a new set of trials on the Cribier-Edwards valve got under way that illustrates one path of convergence between heart surgery and interventional cardiology. The FDA authorized a preliminary set of twenty “transapical” Cribier-Edwards placements. The principal investigators, effectively the trial managers, are Marty Leon and Craig Smith, and the procedures will be carried out at Columbia-Presbyterian, the Cleveland Clinic, and Medical City Dallas Hospital. In a transapical placement, the surgeon makes a mini-incision in the lower chest, places a purse-string suture* in the apex, or bottom point, of the heart, and inserts a catheter. The interventionalist then threads the catheter through the left ventricle to the aortic valve, performs the balloon valvuloplasty, and places the new valve.

The transapical procedure potentially offers multiple advantages. The access route is much shorter, with no concerns over access-route diameters. Unlike the procedure I saw, in which the valve was inserted through an artery to the aorta, or against the direction of the blood flow, the transapical approach follows the blood flow through the ventricle and enters the valve from the bottom, so there should be fewer problems with flow turbulence and less chance of getting stuck on a partially open valve. The first two placements at Columbia-Presbyterian went smoothly. They took place in the new hybrid OR. Mat Williams was the interventionalist and Allan Stewart the surgeon, with Smith, Moses, and Leon all in attendance.

Managing Convergence

In a rational world, professional guilds would organize around diseases and treatments, not around methods of access. But when the favored method of access is the full sternotomy—which is a bit like getting hit by a cruise missile—it will tend to dominate the conversation. But as surgery becomes less traumatic, and interventional cardiology more aggressive, boundaries based on access modes should diminish. Mat Williams’s joint appointment harbingers a new breed of cardiac specialists who can offer patients a menu of choices without regard to guild distinctions. A conference on valve repair would draw valve specialists who offer the full range of techniques from tool-boxes we now label “surgical” or “interventionalist.”

When I spoke to Williams about convergence issues, he made the important point that different approaches come with different success standards. “If you do a transcatheter mitral repair,” he said, “you might be satisfied with a two-plus regurge [an index of valve leakage]. At that level, a lot of people are asymptomatic. But you would never dream of ending an open-chest procedure with a valve like that. If a patient goes through a chest opening and a bypass, he has a right to a regurgitation of zero.” That suggests new possibilities in adapting solutions to patient needs. For many, an adequate, but imperfect, result from a low-stress procedure will be a best-case outcome. When it’s not, you can move on to more invasive approaches. “The important thing,” Williams said, “is not to destroy the bridge to surgery. If your simple procedure would make it impossible to repair the valve surgically later on, then you shouldn’t do it.”

The likely migration path toward the cardiac-interventionalist specialty will probably be from the surgical ranks. For one thing, that’s where the manpower surpluses are, but the surgical craft disciplines also take longer to learn. “The general surgery residency lets you develop skills working on low-pressure operations like hernias,” Williams said, “but there is no easy cardiac surgery.” When the Cribier-Edwards valve is inserted through the femoral, he went on, you need a surgeon to close the wound, since it’s bigger than the one an ordinary cath needle makes. “It’s a simple procedure,” he said, “and a cardiologist could learn how to do it, but it wouldn’t make much sense. If something went wrong, he’d suddenly be facing a major artery repair and would be in very deep trouble.”

However smoothly the reconfiguring of the profession proceeds over the long term, the near-and medium-term challenges will be very daunting. Leading-edge heart centers like Columbia-Presbyterian have been able to maintain their case volumes by radically diversifying their practices and developing a steady stream of new techniques and procedures—aortic and mitral valve repair, valve-sparing aortic root replacement, a variety of new approaches to aneurysms and to fibrillation, new heart-assist devices, breakthroughs in congenital surgery, and much more. But as the kit bag of surgical interventions expands, only the largest centers can maintain adequate volumes in a good cross-section of interventions. Even now, many surgeons in community practice rarely see a valve case. Consider some of the implications:

The large centers will continue to take case share, while medium and small centers will lose volumes. Skills and safety will deterioriate. Should they be closed down?

As programs fall away, will large swathes of the country be left without cardiac surgeons to intervene in heart attacks? And who will provide the surgical backup for the interventional cardiologists?

If residencies are curtailed—either by policy or by lack of interest—how will the teaching hospitals make up the staff losses?

And what will happen ten to fifteen years from now, when the baby boomers hit their seventies and start needing a lot more heart surgery?

These are very difficult issues, of the kind that the professional societies and accreditation bodies are not well equipped to deal with. The natural leadership of the profession, moreover, is centered in the big universities, and is remote from the problems of community surgeons. It is a profession, in short, that is facing interesting times.