Thank you, George, for that very nice introduction.

Ladies and gentlemen.

George, when you called me several months ago to invite me to give this Third Ralph Alley Lecture, I was surprised and highly honored by your invitation. As we talked in that original phone conversation, I had a number of pleasurable memories from more than 40 years of conversations and meetings with Ralph Alley. Foremost among those memories was my very first meeting with Ralph in 1951 when this picture was taken when I was a third-year student at Yale Medical School. Ralph was Gus Lindscog's senior thoracic resident and Harvey Kausel was his junior resident and, for a shy kid from Kansas, these two residents were pretty imposing. I subsequently had many occasions to visit and converse with Ralph over the years and I'll elaborate on those in a moment.

Ralph Alley was truly a remarkable individual and I would like to take two or three minutes at this time to highlight some of the more important events in his productive and colorful career.

Ralph was born in Bulsar India in 1918 to missionary parents. He came to the States at age 18 and attended the University of Virginia and then Yale Medical School. He graduated from Yale in 1943 and after six months as a surgical intern at the Yale-New Haven Hospital, he entered the United States Navy to serve as a medical officer aboard an LST during the Normandy invasion at Omaha beach. This photo of Ralph was taken in southern England shortly before the invasion. After the war, he returned to New Haven to complete his general and thoracic surgical training under Dr. Lindscog's and, after completing his training in 1951, he joined Alan Stranahan's division of thoracic surgery at the Albany Medical College. He had a very productive academic career in Albany and became head of the division of thoracic surgery in 1973. Apropos our discussion on Marfan aneurysm this morning, I should point out that one of Ralph's main clinical interests was surgery for thoracic aneurysm. Here he is, in 1953, harvesting a bovine aorta to use in a patient with an arch aneurysm.

The most important event in Ralph's life was his marriage to a lovely young Yale nursing grad, Jane de Rochemont. Ralph and Jane met at Yale-New Haven Hospital when both were in training. When they married in 1956, Jane was Robert Gross's senior OR nurse. I can't imagine any other cardiac surgeon in the world who has had as many friends and acquaintances as Ralph Alley. People who probably knew him better than anyone, other than his wife and immediate family, are Martin McKneally and Bob Replogle.

Following Ralph's death in September 1996, Martin wrote a moving remembrance of Ralph, published in the January 1997 Bulletin of the American College of Surgeons. I'd like to read just a few of Martin's comments. "Of all of Ralph's roles and contributions, I believe he enjoyed most of all leading his students and residents and providing them with sage advice and counseling regarding their medical careers. Because Ralph was so fascinated with people, he knew the names and life stories of thousands of patients, residents, students, secretaries, orderlies, floor cleaners and elevator operators. When asked about his interests outside of surgery, Ralph answered, 'I have no hobbies, I just collect friends.'"

Martin McKneally also noted in his article that Ralph's most lasting memorial is the Alley-Sheraton Education Fund of the Thoracic Surgery Foundation which, as you all know, provides scholarships to surgeons to develop their skills in public policy and management at the Kennedy School of Government at Harvard.

Finally, Martin related this brief story in his tribute to Ralph. He felt that this was a classic illustration of Ralph's unique energy, magnetism and mentoring. In 1953, Ralph drove on weekends from Albany to the U.S. Naval Hospital at Bethesda to perfect his technique of the aortic arch replacement in the surgical laboratory. There, Ralph recognized a surgical skill and native intelligence of a young Navy corpsman who assisted him in the lab. Ralph steered Seaman First Class Bob Replogle into college, on to Harvard Medical School and to a surgical residency with Bob Gross. Martin continues: one of Ralph's proudest moments came when his protégé, Bob Replogle became the president of an organization that he helped to found: The Society of Thoracic Surgeons. And Bob tells me that Ralph gave him his train ticket from Iowa to Boston for his Harvard Medical School interview. Bob was welcomed into the Alley family like a son. This slide was taken by Bob in 1963 off of Marblehead and this boat had belonged to General George Patton. When he died in Europe, Patton's brother-in-law (Fred Ayer, in the center), inherited it.

Most of you will recognize Ayer's two friends, Ralph Alley and Max Chamberlain.

This Society has three named lectureships honoring Herb Sloan, Max Chamberlain, and Ralph Alley and this remarkable photo by Bob Replogle 35 years ago captured two of these men.

We heard the Chamberlain Memorial paper yesterday morning and I am pleased to present the Alley Lecture at this time.

I have titled this lecture "Antoine Marfan and his Syndrome—One Hundred Years Later."

This was Renoir's Paris in 1888 when Antoine Marfan commenced his medical education at the University of Paris. Marfan came to Paris from Toulouse where he was born in 1858. And this is Marfan as an intern in Paris at age 34. Following his internship, he obtained additional training in pediatrics. In 1896, Dr., Marfan presented this five-year old girl to a monthly meeting of the Medical Society of Paris. This child, named Gabrielle, was unique in that she had disproportionately long limbs that Marfan termed "arachnid-like" or "spider-legs." This case history was published in the Bulletin of the Medical Society of Paris in 1896. And this is that Bulletin, published a little over 100 years ago.

Marfan went on to have an illustrious career, becoming the first Professor of Pediatrics in France and co-developer of the BCG vaccine. It took almost 50 years to fully elucidate the Marfan Syndrome as we know it today. As I said, Dr. Marfan described the arachnodactyly in 1896; dislocated lenses were described in 1914; and the dominant inherited trait in 1931. Then, in 1943, Helen Taussig published a paper in Circulation reporting two teen-aged girls at Johns Hopkins with these three Marfan traits and both had died of a ruptured aneurysm of the ascending aorta. So, it is interesting to note that the most life-threatening component of the syndrome, aortic aneurysm, was not discovered until 50 years after Marfan's original paper.

This is the man who has done as much as anyone to advance our knowledge of the clinical and hereditary aspects of the Marfan Syndrome. This is Dr. Victor McKusick, Professor of Genetics in Medicine at Johns Hopkins, who established the Marfan clinic at our hospital more than 50 years ago. Dr. McKusick is now 75 and, as one of the busiest clinicians at our hospital, continues to see Marfan patients every week in the clinic. Those of you here in the audience today who were doing cardiac surgery in the 1960s would probably agree with me that dealing with an 8-centimeter Marfan aneurysm of the ascending aorta at that time was about the most challenging operation we could do. And here is the title page of Dr. McKusick's textbook "Heritable Disorders of the Connective Tissue," which was published in 1972. And in this textbook, there is a 160-page chapter on Marfan disease and only one page was devoted to surgical management. This was because the early surgical results with large aneurysms such as this using a super coronary sleeve graft. . .the results were miserable. In the 1960s, these patients were never operated upon electively; we always waited until they dissected or ruptured because the operative results were so poor, even for elective surgery.

There is some suspicion among medical geneticists that Abraham Lincoln had the Marfan syndrome. He certainly was tall and had long extremities but he did live well past the average Marfan life expectancy of 32 years.

Dr. McKusick heads a national Science Foundation panel assigned to determine if Lincoln had the Marfan syndrome. Fragments of Lincoln's skull, as well as a blood-stained sleeve of the autopsy surgeon, Edward Curtis, had been preserved at the Walter Reed Museum for Health and Medicine, and this is a slide that Dr. McKusick has taken of these items. Unfortunately, you see only a small area of the blood-stained sleeve over here on the right. These skull fragments and the bloodstain will be analyzed by McKusick's team in the next year or so when DNA and genetic lab

techniques are refined enough to provide a reliable answer about Lincoln and the Marfan Syndrome.

The diagnosis of Marfan aneurysm has always been easy to make if aneurysm is in the differential. The aortogram is classic and of course, the MRI could provide a beautiful demonstration of the aneurysm. This is an MRI of a 9-year old patient of ours and you can see this is a 9-centimeter aneurysm extending from the sternum to the vertebral column.

The very discouraging outlook for Marfan patients in the 1960s changed overnight in 1968 when Hugh Bentall (shown here giving a lecture a couple of years ago at Johns Hopkins) conceived a new operation for aortic root replacement at the operating table. This was the illustration from Bentall’s paper. Bentall was operating on a 19-year old woman with a 9-centimeter Marfan aneurysm of the ascending aorta. There was no way he could place a sleeve graft above the coronary arteries and so he sewed a prosthetic ball valve into the bottom of this tube graft and then lowered it into position and constructed the direct anastomosis to the coronary ostea. With this operation, the surgical results dramatically improved overnight.

You've all seen these next three slides. They are watercolor illustrations—very excellent illustrations—adapted from a paper by Nick Kouchoukos showing the classic procedure and the direct anastomosis of the left coronary ostium and the distal anastomosis.

Over the last few years, I believe most surgeons have adopted the coronary button technique depicted in these new illustrations by our medical illustrator. The button technique allows the surgeon to operate fairly easily on aneurysms that are 5 to 5-1/2 centimeters, even when the coronaries have not migrated significantly from the aortic annulus. There has also been considerable interest, in recent years, in the valve sparing procedures first performed by Ake Senning 29 years ago and reintroduced by Yacoub in 1982. In this operation, the aneurysm is resected, the aortic valve remains intact and the coronaries are reimplanted. Tirone David introduced his version of this operation, called the "David One" in 1992. In this operation, a Dacron graft is slid down over the aortic root. More recently, Tirone introduced the "David Two," which is somewhat like Yacoub; but he utilizes a Teflon felt strip to prevent annular dilatation.

We performed our first vental procedure at Johns Hopkins in September of 1976 and, over the last 21 years, we have replaced the aortic root in 231 Marfan patients. Two hundred eighteen had composite graft repairs; 11 had aortic root homograft repair; 2 had the valve-sparing procedure. You will note that there were 168 males and 63 females. One hundred and fifty of these 231 patients were mine and the remaining 81 patients were operated on by 10 other Hopkins surgeons. The average age at the time of surgery was 32.8 years, with the age range from 4 to 73 years.

Here we depict the aortic root diameter of the sinus level in 223 adult patients. You will note that the average diameter was 6.8 centimeters, the diameter range from 4.5 to 10 centimeters. The average diameter of 43 dissected aortas was 7-1/2 centimeters but, most importantly, 14 patients had dissection with an aortic diameter of 6-1/2 centimeters or less, thus emphasizing the point that Marfan aneurysm should be resected by the time they reach 5-1/2 and certainly 6 centimeters even if the patient is asymptomatic.

Here are the operative results in the 231 Marfan patients with aortic root replacement at Johns Hopkins: 198 had elective repair; we were very fortunate to have no 30-day mortality in this group. Thirty-three had urgent repair and there were two deaths in this group. Both these patients who died arrived in the operating room with frank rupture of the ascending aorta and were moribund before surgery was initiated, so the overall 30-day mortality was 0.9%. Morbidity was in three main areas: endocarditis, thromboembolism, and late coronary anastomotic dehiscence. The most serious late complication was endocarditis. We had 8 patients with late endocarditis; 2 that had a very mild infected composite graft could be cured with antibiotics, but 3 that had severe infection of their composites could only be cured with a homograft root. There were 3 late deaths in this group of 8 patients and all these patients died without benefit of a preserved homograft; we principally used another composite graft which, in our estimation, simply doesn't do the job. These patients, if they have a significant infection of their composite graft, need a homograft. And the other two complications, I say late complications, there were 7 patients with thromboembolism and 4 with late coronary dehiscence. Actually, these 7 patients with thromboembolism: one patient had a partial thrombosis of the Bjork Shiley valve which successfully replaced at 10 years. There were 6 cerebral emboli and these were all. . .5 of them were early: 1, 3, 12, 14, and 35 days—probably inadequate anticoagulation; one at 9 years; all with complete recovery. This is about the instance of embolization with a prosthetic valve alone, which may relate to the difference in the valvular suture line. And there were 4 late coronary dehiscences: 3 successfully repaired at 1 month,2 years, and 13 years; 1 sudden death at 7 years post-op.

Here we depict the actuarial survival for the 231 Marfan patients; there was 88% survival at 5 years; 81% at 10 years; and 75% at 20 years. Looking at the univariate and multivariate risk factors for mortality, you will note that the significant risk factors were poor New York heart classification, that is, class III and class IV, male gender and urgent surgery. Preoperative dissection, age, and concomitant mitral valve surgery were not risk factors for mortality.

I would like to present just one more data slide. Remember that Antoine Marfan presented his 5-year old patient with long extremities in 1896. To commemorate the centennial of Marfan's landmark paper, the International Marfan Foundation organized a symposium in Davos Switzerland in the summer of 1996. I was asked to summarize the surgical results from several major Marfan centers around the world for this meeting and we obtained data on 675 Marfan patients from these 10 centers. The centers and the principal cardiac surgeons are listed and I will take just a minute to let you note the centers and the senior surgeons from these 10 centers. I believe that 8 of these 10 surgeons happen to be at this meeting and may be here today.

If we look at the operative results of these 675 Marfan patients with aortic root replacement (incidentally, these are all of the Marfan cases; none have been excluded from these centers with the exception of Methodist Hospital. We have all the cases from Dr. Coselli and his associates; we don't have the cases prior to Dr. Coselli). But every other center, we have all the Marfan cases going back 20 years. If we look at these operative results of these 675 Marfan patients with aortic root replacement, we will note that for the 107 patients who had emergent repair, that is, surgery was mandated within 24 hours, usually because of a dissection, there was an 11.2%, 30-day mortality. If urgent root repair was carried out within the first 7 days of admission, the 30-day mortality rate dropped to only 2.6%.

I think this is a remarkable figure down here. For 454 patients undergoing elective repair, the 30-day mortality rate was only 1.4%. These are the results from 10 different centers going back over 20 years and I think the results are pretty remarkable.

During the period of this study, 50 of these 675 patients underwent a valve-sparing procedure with excellent results. I believe, however, that we need to use the valve-sparing operation in Marfan patients with a certain degree of caution because there is some evidence that some of the same structural deterioration observed in the aortic wall is present in aortic leaflets. Recent histologic studies at our institution, primarily by one of our cardiac surgical residents, Curt Fleischer, demonstrated a surprising amount of structural deterioration of aortic leaflets excised from Marfan patients. Here, for example, here's an example of the histologic structure of a normal aortic leaflet and here is the abnormal structure—from here it doesn't project well at all—but here is the abnormal structure of a Marfan aortic leaflet. An immunofluorescent staining technique has been used and it can be seen that the fibrillin in a Marfan leaflet is markedly fragmented. So these recent studies showing fragmented fibrillin in all excised Marfan leaflets would suggest that caution be used in the routine use of valve-sparing procedures for Marfan patients. We feel at the present time the operation of choice in a Marfan patient is still the modified vental procedure but the valve-sparing operation probably has a place for younger patients and for women who would like to carry a successful pregnancy.

Clearly the outlook for Marfan patients undergoing surgical repair of the ascending aorta is very gratifying. Accompanying the steadily improving surgical results in Marfan patients, there have been spectacular developments in understanding the genetic picture in these Marfan families. In attempting to be fairly complete in discussing the Marfan syndrome, I would like to devote the last five minutes of this presentation to some recent genetic discoveries—many by Johns Hopkins scientists. It has been known for some time that connective tissue of Marfan patients contain defective fibrillin. This not only explains the abnormal findings the excised aortic valve leaflets but also in the aortic wall. Defective fibrillin can account for these scanning electron micrograph studies published 10 years ago by Pevejda. The staining here is for the other major protein and connective tissue elastin. In fact, scientists have only been able to stain for fibrillin for the last 7 or 8 years but have been able to stain for elastin for many years. So this is an elastin stain and these studies show that in the normal aortic wall, there are these longitudinal parallel lamina of elastin and there, if you'll look carefully, you'll see perpendicular strands of elastin that maintain the integrity of this aortic wall. Here are three scanning electronmicrographs from three different Marfan patients and you will notice that, in all three, there tends to be an absence of these perpendicular elastin strands with actually a Swiss-cheese-like appearance.

We've been talking about these two proteins, elastin and fibrillin that provide strength to connective tissue and for some time, I've been trying to come up with a recognizable model to depict these two proteins. A couple of weeks ago, I was passing this construction site at Johns Hopkins stadium and I thought what a good analogy this column provides for these two proteins. The main strength of this column comes from the steel reinforcing rods but added strength comes from the surrounding concrete. I think this structural column serves as a good model for the fibrillin rods shown here and the encompassing elastin in the connective tissue.

In 1990, a Finnish research team led by Katarina Kainulainen demonstrated that the fibrillin gene is located on the long arm at chromosome 15. Then, in 1991, Dr. Harold Dietz and his associates at Johns Hopkins reported their discovery of the first mutations of the fibrillin gene. This is an illustration taken from Dietz's publication. It shows that the fibrillin monomer is made up of these very precisely pleated motifs, as you see here, each one with 42 amino acids. These pleated structures are strung together, as you see, and they make up two-thirds of the fibrillin monomer. It's essential that fibrillin be able to bind calcium and this pleated structure is necessary for this property. And critical to maintaining this pleated configuration are these six cystine amino acids that are bonded together by disulfide bonds—these are the bonds that you see here. These cystines are critical to this pleated structure.

As you may note on this slide, Dietz's first four (four of his first mutations that he discovered), affected these cystine amino acids. Here, for example, a serine has replaced this cystine because of a mutation at. . .I can't read that; it looks like the 351st nucleotide of the fibrillin gene and so, a serine has replaced that. Over here, another serine has replaced this, an arginine has replaced it and so far now, there have been about 75 different mutations noted, most of them affecting (substituting) a different amino acid for the cystine. So if the calcium-binding properties of fibrillin depend on this pleated structure, and the cystines are replaced by other amino acids, this fibrillin structure could come apart like a spider web in a windstorm.

This inability of mutant fibrillin to properly bind calcium in Marfan patients is dramatically depicted in this slide originally published by Godfrey and his associates at the University of Nebraska. This figure shows immunofluorescent-stained fibers of fibrillin in tissue cultures from a family with the Marfan syndrome. The tissue culture of the unaffected father is shown here and his unaffected daughter is shown here. And this is a tissue culture of the affected mother and two affected sons, shown here.

Dietz and his associates have attempted, over the last three years, to develop a mouse model with the Marfan syndrome. These investigators have injected abnormal fibrillin genes from Marfan patients into single fertilized mouse ova and have created a mouse with the Marfan syndrome. Here is one of Dietz's normal mice and here is a Marfan mouse with a kyphosis and long. . .it's got long feet as you can see in this slide. Most of these Marfan mice develop aneurysms such as this, looks just like a human Marfan aneurysm. This aneurysm has dissected and ruptured. Most of these Marfan mice will develop this kind of an aneurysm of the ascending aorta, dissect, and rupture. Dietz is now working on a pig model of the Marfan syndrome.

The ultimate hope of Marfan families is to eliminate the disease completely; however, this seems virtually impossible, at least within the next 20 years.

It was at the turn of the century that Sir William Osler stated, "There is no disease more conducive to clinical humility than aneurysm of the aorta." That was certainly true in 1896 when Antoine Marfan described his 5-year old patient at that Paris medical conference. Fortunately, we now have an operation for Marfan patients that provides excellent long-term results. Unfortunately, hundreds of Marfan patients with life-threatening aneurysms will go undiagnosed every year, even when these patients present to the emergency room with classical signs and symptoms of acute dissection. Let me just share a couple of examples of misdiagnoses with you. Here for example, is America's number-one volleyball player, Flo Hyman, shown in the middle. This photo was taken in Tokyo at the 1984 Olympics when the American team won a silver medal. Flo Hyman was, at the time, 29 years of age and 6’6" tall and nobody suspected the Marfan syndrome. She died of a ruptured aorta during a volleyball game in 1986. And here is a 25 year-old young man who came to Hopkins for repair of his Marfan aneurysm. He had been a high-school and college basketball star. He signed with the Philadelphia 76ers. He played for 4 years in Japan on their farm team but when he returned to Philadelphia to play for the 76ers, a routine echocardiogram was obtained. And he had a 7-centimeter aneurysm. This man was 6’10" tall. He played all the way through high school and college and four years of semi-pro ball but nobody ever suspected the Marfans.

And, finally, the recent and tragic case of Jonathan Larson. Larson, as you know, conceived and produced the musical "Rent." Three days before the show opened, he had severe chest pain and was seen at an emergency room in a New York City hospital. Chest X-rays were obtained and he was sent home with a diagnosis of gastritis. The next day the symptoms reappeared. He returned to another emergency room, again with excruciating chest pain. A chest X-ray was obtained and, again, he was sent home with a diagnosis of gastritis. tragically, he died a few hours later, on the evening before his musical "Rent" opened on Broadway. He had ruptured a large Marfanoid aneurysm of the ascending aorta and the New York Department of Health felt that the misdiagnoses in these two emergency rooms were such egregious events that they fined both hospitals several thousand dollars.

It's tragic that, in this day and age, patients such as Jonathan Larson can present to an emergency room with sign of aortic dissection and a suspicious chest X-ray and not have the correct diagnosis made.

In conclusion, we can say that we now have an exceptional operation for this problem that was so difficult to manage 25 years ago. Marfan aneurysm patients can undergo elective surgery with an operative risk well below 5% and have an excellent likelihood of normal life expectancy. It's essential, though, that a timely diagnosis be made by our non-surgical colleagues.

It's been a great privilege for me to give this Third Ralph Alley Lecture. We all owe so much to Ralph, not only for his surgical and scientific achievements, but also for his contributions as a remarkable human being.

Thank you very much.