Marcel Kerboull has had a profound influence on the development of hip replacement surgery in France. His management principles, his meticulous surgical technique, and his great clinical experience have made him a leader in his discipline. Meet the master through this interview with Maîtrise Orthopédique.

 

Kerboull M.O. Let’s start with the usual question: Why did you become a surgeon?
M.K. I would say that I became a surgeon more by chance than by design. I certainly went to medical school by chance, because I had not intended to be a doctor. I was much keener on maths; and because I didn’t know what to do in my native Brittany, I went to spend my holidays in Paris, where I had relatives. On the train, I met a fellow student from my school, who was going to read Medicine. I did not take a decision right there and then, but I thought about it, and in the end I enrolled at medical school as well.

M.O. Was there nobody whose example you were following?
M.K. I remember that when I was little - at the age of 10 or so - I wanted to be a doctor, because I had an uncle who was one, and I thought that was super; but then I gave up the idea. When I decided to enrol at medical school in Paris, I was told that I couldn’t, because I was not resident there, and that I should go back to Rennes. I was going to, but got on the wrong train, and went to Angers instead of Rennes. I got off at Angers. I hadn’t been there before, but I liked the town. I spent two days in Angers, found the place pleasant, discovered that they had a medical school, and went to enrol there.

M.O. Perhaps you had a subconscious urge to get away from your family?
M.K. Very likely.

M.O. When were you at Angers?
M.K. In 1952. I had no problem with the first part of the exams, the science subjects, because I had had a good grounding in maths, physics, and chemistry. So I started my first year at medical school, and then sat the exam required for the clinical part of my training at Angers. I passed easily, as did the handful of other undergraduates who were taking the exam at the same time. However, I was still tempted to go to Paris, so I worked for the Paris exam.

M.O. Could you do that?
M.K. In those days, yes, one could. One would spend three months in Paris, in a hall of residence, and attend lectures at the hospital, or in what was then known as the Student’s House - a centre opposite the Cochin, which had been built at the end of the war, with American money. I got my clinical place in Paris quite soon, and left Angers in my third year.

M.O. Had you already developed a liking for like surgery?
M.K. Initially, I was not at all drawn towards surgery. I had done a six-month stint at Angers. In Paris, the clinical training stage started off with a compulsory year of surgery, and I decided to go to the Cochin, to Merle d’Aubigné’s department, because I had heard that it was a good place, where one learnt a lot.

M.O. And that sealed your fate?
M.K. With hindsight, one could say it did. What happened was that I spent a year in Merle d’Aubigné’s department, and learnt a lot. There are things that were “ground into me,” as the chief would say - things that became part of me, although I couldn’t say offhand what sort of things they were. In actual fact, I left Surgery after that one year, and was only too happy to turn to Medicine.

M.O. With Merle d’Aubigné, the clinical student was the fourth wheel of the car?
M.K. Yes - or, perhaps, the fifth: the department could function perfectly well without any clinical students. We were mainly required to assist at operations, but we were full members of the team. Each clinical student was responsible for a certain number of patients on a ward (hospitals still had large wards in those days). They had to clerk them, and do the obs after surgery. Every now and again, they had to present a clinical case to the Saturday morning conference attended by the Chief and all the medics in the Department. Merle d’Aubigné was beginning to mellow a bit by then, but everyone still stood in awe of him. You could even say that he put the fear of God into people. However, he had instituted a way of working that I found immediately attractive, because everything was perfectly organized, with proper schedules for what we had to do mornings, afternoons, and sometimes evenings; and the clinical students were fully integrated into the system.

M.O. So, you then went and did Medicine....
M.K. I thought that Medicine was more rewarding, and required more of an intellectual effort. Orthopaedic surgery had struck me as a craft, rather like furniture-making or car repair: basically, it was a technique. At least, that’s what it looked like to a clinical student who had been on an orthopaedic placement. Medicine, on the other hand, - well, one need only look at the medical questions in the entrance exam for specialty training, as opposed to the surgical ones, to see how much richer and more varied the subject matter is, and how much more uncertain everything is. And these were the things that I found attractive.

M.O. So you started your specialty training....
M.K. Yes. I failed the entrance exam first time round, and passed the second time. In between, I had spent some time in other departments on the medical side, including six months in Paediatrics. However, the way I remember my clinical student days on the medical side, there was a lot of wasted time, and one was not really doing anything to help the patients, especially in neurology and cardiology. In cardiology, the clinical student’s first job, at 8 o’clock every morning, was to go to the morgue and get the hearts of the patients that had died overnight. These hearts were then examined with the Chief. This way we could see the pathology that we had been listening to the night before. That, obviously, is a useful way to learn cardiac pathology and anatomy. However, I wanted to do more; so when I had passed the specialty training entrance exam, I opted for Surgery.

M.O. Had you, by then, decided to go in for orthopaedic surgery?
M.K. When I started training as a surgeon, no. I first did obstetrics, because in central Paris, in one’s first six months, it was difficult to find a placement other than in Obs, ENT, or Urology. I chose obstetrics, because I was quite intrigued by the mechanics of delivery; and Obs went hand in hand with Gynae, with its medical content. This was a transition, as it were, from the Medicine that I was leaving behind to the Surgery to which I was turning. In actual fact, I did not do much obstetrics, because I had to take 4 months out to have surgery. I had, for some years, been suffering from bilateral recurrent pneumothorax; fortunately, it never recurred simultaneously on both sides. Two years before, I had been treated with intrapleural aureomycin injections. These injections are irritating, and had caused some adhesive bands. During a subsequent recurrence, one band, which contained an arteriole, tore. This produced a haemothorax of 2 litres within a few hours, and I had to be operated on as an emergency, by Claude Dubost.

M.O. Where did you go then?
M.K. I went to Moulonguet, who ran the big Department of General Surgery at the Salpêtrière. He was a lovely person, very, very nice and pleasant. He had a great sense of humour, and a very wide general education, which I, for one, found impressive. I also met Loygue there.

M.O. And did you like general surgery?
M.K. Yes, I liked Loygue’s way of doing surgery. He did lots of colons, and everything seemed easy when he did it.

M.O. Perhaps because of good patient selection?
M.K. No, that was not the reason at all. He may have had some easy cases, but most of the patients had been operated on before. I always admired the elegance, the restraint, and, above all, the effectiveness of his surgery. He was always perfectly in control; he also had something that I had already seen with Merle d’Aubigné: he knew that a surgical procedure is something that has to be thought out well in advance, so as to protect oneself against surprises at surgery, or at least to foresee what might happen and how to cope. It was one of the things that I had learnt as a clinical student, and that I came across again in my specialty training.

M.O. Where was your next placement?
M.K. With Sicard. However, I had to stop after two months, because I was continuing to have pneumothoraces on the non-operated side. Claude Dubost operated on me once more.

M.O. Did this interfere with your specialty training?
M.K. Yes, I had to stop twice for 4 months. In actual fact, I only took twice 2 months out, because I spent half of my sick leave assisting Gouzon, who was also a very skilful surgeon and whom I admired greatly. He worked with Chigot, who was at a private clinic. He did both paediatric and general surgery. I learnt a lot from assisting him three or four times a week. I then went back to my training post with Sicard, because I was interested in the vascular surgery done by Natali. Vascular surgery, in those days, was pretty dicey, but fascinating. In the laboratory at the Salpêtrière, I tried to do some experimental vascular surgery in dogs. There was one particular problem at the time that had not been resolved: uncontrollable arterial spasm. Nowadays, I think, there are drugs that can be used. I never managed to induce that arterial spasm in the dogs. However, after this first study, this complete flop, I had at least acquired some experience in vascular surgery, and I had done a comparative study of vein grafts versus Dacron arterial prostheses. This study, which took over a year and for which I had to sacrifice a large number of dogs, only confirmed things that had been well known before. I still feel guilty about having sacrificed all these dogs for nothing.

M.O. But being able to repair arteries was at least an interesting subject?
M.K. Oh yes, indeed. After that, I went to Chigot at the Trousseau. He had an enormous department, with general surgery, orthopaedic surgery, and traumatology wards; he also dealt with all manner of malformations, and had a burns unit. There was also a small paediatric urology ward under Bruzières, another surgeon who interested me greatly and from whom I learnt a lot about urology. I found everything fascinating. I was also still working with Gouzon, who did mainly neonatal surgery, especially for atresia of the oesophagus.

M.O. You did the procedures yourself?
M.K. Very few; but I assisted the surgeons, on many occasions.

M.O. But you would do some surgery when you were on call?
M.K. Some, yes. In Sicard’s department, emergency general surgery other than appendectomies would be done by the Chief himself. Traumatology cases were generally left to the juniors. Many fractures were treated conservatively, even proximal femoral fractures. With Chigot, on the other hand, there was always plenty to do when on call, and the juniors were kept pretty busy.

M.O. Then, after Chigot - where did you go?
M.K. I went to Merle d’Aubigné, for my third year of specialty training.

M.O. What was it like, going back?
M.K. The big new Ollier unit had just been opened, and it was palatial. Everything was brand new, everything was vast. From the basement to the fifth floor, everything was laid on: X-rays, labs, physiotherapy, superb teaching facilities. This really was a “fully seaworthy” flagship, as Merle d’Aubigné used to say. Everything was working even more efficiently than before, because the system he had set up and the facilities that he now had were more suitable. However, he was starting to delegate some of his work to his senior surgeons - Ramadier, Meary, Postel, Maurer, and Tubiana. Even so, he would still see, on Saturday mornings, all the patients scheduled for surgery the following week. These patients would have been admitted on the Thursday or Friday. They were examined by the registrar, the senior registrar, the staff consultant, and, finally, presented to the Chief himself on the Saturday morning. They were due to have their ops the following week. For those on the Monday list, this was bearable; however, for those who were not due to be operated on until the Friday, this meant a preoperative in-patient stay of eight days.

M.O. Were you mainly doing hips, by then?
M.K. No, we had a wider spectrum than the one in my present-day department. However, there were lots of hips, lots of congenital malformations, also lots of primary OA; above all, there was something that we are seeing less nowadays: residuals of trauma.

M.O. Was it a good department for a junior surgeon?
M.K. Yes. I liked the strict logic applied to each case, the search for the best treatment modality for each patient, so as to produce the best possible outcome. Of course, in those days, the outcome of joint surgery was much less glorious than it is today.

M.O. How was treatment decided upon?
M.K. It was the task of the registrar, the senior registrar, and, where appropriate, the senior surgeon, to suggest a treatment, to defend that proposal, and to list the other options. In the end, the decision would be taken by Merle d’Aubigné. The clinical conferences were mainly teaching sessions. However, while this teaching was very instructive, it was not always pleasant, because the seniors were often very harsh with the trainees.

M.O. Were there any total hip replacements at that time?
M.K. No, but there was the Moore endoprosthesis. As a clinical undergraduate, I had seen acrylic implants, but they had disappeared in the meantime. A Garden IV femoral neck fracture in an old lady would be managed with an endoprosthesis; however, for a Garden III, a Smith-Petersen nail would be used.

M.O. What were you allowed to do as a trainee?
M.K. There were seven or eight junior surgeons in the department. This meant that each one would be on call once a week, together with a senior registrar, a final-year undergraduate, and an anaesthetist. Through A&E, we got the usual assortment of big city trauma cases, which has not changed that much since those days; we would also get some highway accidents, because the peripheral facilities that would deal with these cases nowadays had not yet been created. So, most times, one would see a patient with severe multiple injuries. These patients would have head injuries, but often also abdominal trauma. The duty team would deal with bony lesions, as well as with visceral trauma, especially injuries to the spleen. Conversely, Léger’s general surgical team also got its share of hip fractures, which were managed with endoprostheses. However, things were beginning to change - although it was only later, when I returned to the Cochin as a senior registrar, that general surgery and orthopaedic surgery ceased to be done by the same surgeons.

M.O. What sort of hardware did you have for fractures?
M.K. We had everything, especially every possible and conceivable Kuntscher nail - for the humerus, the tibia, the femur, the clavicle, the ulna. We did a lot of nailing.

M.O. Open nailing?
M.K. Yes, except in the tibia. The tibia was nailed without reaming. The femur would be nailed open, with reaming. Next morning, we had to show the Chief the notes of all the patients we had operated on while on call; at the same time, the X-rays of patients operated on electively the day before would be shown; and we would be given a rough time if our reduction or internal fixation was not perfect.

M.O. By the end of that year, had you decided on a specialty?
M.K. Not yet. I did my last 6-month placement with Thomeret, and came to like thoracic surgery so much that I stayed on for another 6 months after the end of my specialty training period, on a sessions basis.

M.O. So, by the end of your specialty training, you had been well trained as a general surgeon, and, as a sideline, as an orthopaedic surgeon....
M.K. Yes, that was how surgeons would be trained in those days. In 4 years, I had actually only done 18 months of orthopaedics - 1 year of adult, and 6 months of paediatric surgery. I was undecided at the time whether to go in for paediatric surgery or for orthopaedics. In the end, I opted for orthopaedics, and went back to Merle d’Aubigné, as a senior registrar.

M.O. Why did you do that?
M.K. Because that’s what appealed to me most. I liked his way of working and of thinking. Also, paediatric surgery had not yet been split up into orthopaedic surgery, urology, and everything else. Specialization was beginning to happen, but Chigot did everything. So did Petit, in those days; and I couldn’t see myself doing that.

M.O. So you were a senior registrar at the Cochin. Tell us how, around 1960, a Garden III fracture would have been treated.
M.K. It was reduced on a fracture table, with slight traction and internal rotation. Then X-rays in two planes were taken, to assess the reduction. In very rare cases, a Leadbetter manoeuvre - reduction with the hip flexed - would be used. For this, the patient is positioned supine; with the hip flexed to 90°, traction is applied with the knee flexed to 90°; this is followed by extension in internal rotation. This often reduces a Garden IV fracture.

M.O. After reduction, how would you perform fixation?
M.K. With a Smith-Petersen nail; however, dynamic hip screws were coming in.

M.O. Why not use a large screw?
M.K. The big, 7-mm screws then available were not cannulated. It was easier and safer to put a K-wire into the centre of the head and the neck, and to insert a nail over the wire.

M.O. With a Garden IV fracture, would you use a Moore endoprosthesis?
M.K. In an elderly patient, yes. Incidentally, Moore himself had come to us a couple of years before, to show us how to do his device.

M.O. What can you tell us about Moore?
M.K. Austin Moore was an American. “No touch” was not a concept he was all that familiar with, and we were a bit shocked when we saw him operate.

M.O. And total hip replacement?
M.K. THR came in just as I was starting my senior registrarship under Merle d’Aubigné, in September ‘65. Watson Farrar came and did the first one, a metal-on-metal McKee.

M.O. What did Merle d’Aubigné think of that?
M.K. He had made his choice, after “shopping around” in Britain. He went across the Channel at regular intervals, and was also very Anglophile. The advent of THR had been expected for some time. We had not got one, but we knew that these implants existed. In particular, we knew that two Brits had been working on THR for several years. So Merle d’Aubigné and Postel went over to see how things were coming along. At that time, the McKee was perfect, without any major problems. John Charnley had had a disaster, with a steady stream of patients who had to be revised because of their failed Teflon sockets. He had been experimenting with a PE socket for the past two years, but had not authorized the release of this component.

M.O. What did you think of Farrar’s operation?
M.K. Farrar used an anterolateral Watson Jones approach, with the patient supine. I did not like this technique very much, because one could not see the acetabulum properly, and preparing the femur was difficult. In order to get a proper view, a little too much of the femoral neck had to be resected, which would leave the patient with one leg slightly shorter and a mechanically unsatisfactory hip. Very soon, therefore, with his third THR, Merle d’Aubigné positioned the patient side-lying and used a lateral transgluteal Gibson approach, which is how things had always been done in his Department. This is a lateral approach, which involves the division of the tendons of gluteus medius and minimus 1 cm from their insertions on the trochanter, leaving the short rotators intact.

M.O. Why did you abandon this approach?
M.K. We went on using it for two years. Then, one day in 1967, I was assisting Postel with a THR in a lady who already had a Moore endoprosthesis; and just as he was about to cut the gluteus tendons, I pointed out that there was hole behind, and perhaps one could go through that. So he did, and I think that was the first time that a THR was implanted through a Moore approach.

M.O. Nobody had done that before?
M.K. I don’t think so, but one would have to check. Next day, I did a THR through a Moore approach in a case of primary OA. Very soon, this technique was picked up by all the surgeons in the Department.

M.O. No complications?
M.K. With this approach, we started off a whole series of thromboembolic complications, because we thought that, since the patients were mobilized the day after surgery, there should not be any need for anticoagulation. Within one week, we had three cases of DVT. So all the patients were once again put on anticoagulants; and then we tried to find out why his should have happened, using a Moore approach and femoral vessel opacification in fresh cadavers. We found that the vessels were nipped when 90° internal rotation was combined with adduction and a certain degree of flexion. While we were at it, we tried to see whether a different position, anterior dislocation through the lateral incision, would leave the arteries and veins unobstructed.

M.O. What were the patients like postoperatively when you had cut the gluteus tendons?
M.K. In these cases, the patients would be in traction for a fortnight, same as after mould arthroplasties, where they were sometimes kept in bed for three weeks. In those days, there was not much surgery being done. The mean in-patient stay of a hip case was one month.

M.O. Did your cadaver studies make you abandon the Moore approach?
M.K. No, we did not give it up, because it is a very convenient approach. For a while, we did both Gibson and Moore approaches. Then everything, even revisions, came to be done through a Moore. I think it was in ‘68 that we decided that it would be interesting to compare the different approaches. To make things more interesting, the transtrochanteric approach was also included. We had a large series of Gibson and of Moore approaches, and 200 transtrochanteric ones were added to those.

M.O. Did you publish that study?
M.K. In 1969, we took stock, in my first paper on THR. I had done the usual analysis of the results, and added an assessment of the outcome as a function of the approach used. However, Merle d’Aubigné did not want any mention of technique in the Presse Médicale, where the paper was going to appear, so this aspect of the subject was never published. However, we found that, at 6 weeks, the patients who had been operated on through a Moore approach were doing markedly better than the others; at 6 months, though, the ones operated on using a transtrochanteric approach would have caught up and overtaken the Moore approach patients.

M.O. Why was that?
M.K. There are several reasons. Those operated on through a transtrochanteric approach had better implant alignment. Overall, these patients had a more balanced ROM, and fewer venous or sciatic nerve complications. Also, they had a lower rate of heterotopic ossification. Often, there would be a spot of ossification at the lateral trochanter resection line, but hardly any elsewhere. By contrast, patients who had had a muscle transection - either of the gluteals or of the short rotators - frequently had heterotopic ossification, with foci proximally, in those who had had the gluteals divided; or posteriorly and especially distally, in those who had been operated on through a posterior approach. This gave rise to stiffening, often in a bad position of the hip. However, with the large, 41-mm diameter heads, there was no dislocation. So it was really the poorer function at six months that made us abandon the Moore approach in ‘69; and since then, all our cases have been done by the transtrochanteric route.

M.O. Do you think that a posterior approach really carries such a risk of thrombosis?
M.K. With routine heparinization, this complication has become rare. However, in the different multicentre studies of low molecular weight heparin, the risk appeared to be greater than with a lateral approach.

M.O. The time required for surgery also plays a role. How long did it take in 1968 to put in a THR?
M.K. Not very long - often just under 2 hours.

M.O. How long were you there as a senior registrar?
M.K. Two years. After that, I had some hope to be among one of the last batches of those who could take the tertiary level teaching exam; but it was not to be. Everything came to a halt for 4 years. I therefore went into non-hospital practice, but was still available, virtually full-time, for work at the Cochin.

M.O. You worked as a staff consultant at the Cochin?
M.K. In 1968, Merle d’Aubigné split his big department into two units. He kept one, and Michel Postel took over the other. I found myself on his unit, and was working mainly for him. I also went on assisting him in his private practice, because he had still only a part-time contract with the Cochin.

M.O. When did you stop doing McKee-Farrars?
M.K. In 1969. By then, we had not done many revisions, but I was unpleasantly surprised to see that, after two years, a considerable number of patients had complete demarcation around the socket, on the a.p. and the lateral views, as well as some pain and a slight limp. We thought that this actual or potential loosening of the cup was due to excessive friction. We measured the retrieved implants, and found that the friction torque was, indeed, very high. The low-friction principle was just coming in, and it looked as if the early problems that had been encountered did not exist with PE. So, when Charnley released his LFA in 1969, we used it; however, we could get hold of only two or three of these prostheses. This is why Michel Postel decided to have the implant made in France.

M.O. There were supply problems with the LFA?
M.K. Yes. The same had happened back in 1965, with the McKee, which was one of the reasons why Merle d’Aubigné developed his own device. The other reason was that he did not like the femoral component.

M.O. What did you like about the Charnley prosthesis?
M.K. Its metal-on-PE combination, with very low frictional torque. On the test rig, comparing it with the McKee-Merle d’Aubigné, it was like chalk and cheese.

M.O. Where did the rig come from?
M.K. We had had it built ourselves; it was in a basement lab. At that time, we had an engineer who was very skilful and could put his hand to anything. All our surgical instruments were repaired, and some of them even designed, in-house.

M.O. Isn’t it difficult to redo a Charnley?
M.K. Since we had this problem getting hold of LFAs, we had asked Charnley’s permission to have the implant made in France. He agreed, provided that the French device was called “Charnley-type.” After some trial and error, the French implant finished up looking exactly like the original Charnley.

M.O. So, initially, you kept the same features, the same angle....
M.K. Indeed - both for the standard and for the revision prostheses. However, instead of using the revision device for failed displacement osteotomies, the way Charnley did, we used it in congenital dislocations and their residuals, i.e. in dysplastic femurs.

M.O. And then you went and modified the Charnley. Why?
M.K. Because of something that we started seeing in 1971 - not around the cup, which was fine, but in the femur. There was a very worrying X-ray pattern: a lucency at the convex portion of the implant, often associated with a transverse fracture of the cement mantle at the stem tip. The implant would settle a bit, go very slightly into varus, and then stabilize more or less at the new level. We told Charnley about our concern, and he said, “Yes, that’s common, it’s nothing to worry about. Also, the patients don’t complain.” However, the patients with these abnormal images were not always asymptomatic; and often the lucencies would be seen after as little as 6 months. I tried to investigate this problem, and collated 200 sets of patient notes. I found that, at 2 years, I was having almost as many problems on the femoral side with the Charnley as I had had on the acetabular side with the McKee. I thought this was untenable, because even though many of these loosened implants were not causing any major clinical problems, the eventual outcome would be poor; so something needed to be done. My first idea was to have a hybrid implant, with a Charnley socket, a Charnley head and neck, and a McKee-Merle d’Aubigné stem, whose cemented fixation had proved to be very reliable. I designed this device in 1971, but it never say the light of day, because it was an ugly-looking thing. So I started some brainstorming and some studies. As I saw it, the Charnley stem tended to loosen within the cement mantle for the following reason: The device has a thin stem, with a long abductor lever arm, and a tight neck-shaft angle. This produces unduly high compressive stress on the cement in the superomedial zone, resulting in a single or double vertical fracture on the medial side, which will widen the upper part of the cement mantle. Since the implant is then no longer fixed in its proximal third, it will exert considerable pressure on the cement distal to the tip, which will then fracture transversely. In order to improve the fixation of the implant in its cement mantle, I first thought of increasing the bond between the implant and the cement by giving the device a rough surface. I soon gave up that idea, because I thought that the resultant increased stress at the cement/bone interface would be too risky. I therefore decided to stay with a smooth surface, but to change the shape of the stem. This modification mainly concerned two features: the neck-shaft angle, and the cross-section of the implant. In order to decrease the pressure on the superomedial part of the cement mantle, I reduced the offset and made the neck-shaft angle 130°. By thickening and widening the metaphyseal portion, I obtained a more tapering cross-section. My hope was that the shear stress along the stem would be transformed into its horizontal compressive component, so as to reduce the residual force acting on the most distal part of the cement mantle, and thus to prevent cement fracture at that site.

M.O. That was quite a complicated way of reasoning, when what you really wanted was a thicker stem....
M.K. Complicated - no; elementary is what I would call it. The only reason for having a thicker, and therefore more rigid, stem is because it puts less stress on the cement than does a thinner and more deformable implant. However, a thicker stem will not, by itself, protect the cement against excessive stress levels. I have designed big stems, and medium-sized stems, and small stems; I have had a whole range, with 5 neck lengths and 4 offsets, to cope with all anatomical patterns and to provide the patient, in each case, with an artificial hip that comes as close as possible to the normal anatomy. All these stems were mechanically alike, in that they reduced the adverse tensile and bending stresses on the cement, and produced compressive stress, which the cement can cope with much better. However, for the cement to stand this stress, the host bone must be smooth and rigid - in other words, you need cortical rather than cancellous, bone, which is why it is so important to remove all cancellous bone from the canal. This applies not only to the superomedial and inferolateral zones, where the cement is compressively stressed by the way the implant is shaped, but all along the stem, because of the way that shear stress is transformed into horizontal stress, and because of the axial torsion forces acting on the stem, which transforms torquing into compressive stress.

M.O. You were getting quite close to the theory of cementless fixation.
M.K. I even came close to giving up cementing in 1971, when we were having all this trouble with the McKee and the Charnley. I designed a range of cementless implants then, which would not look out of place among the present-day patterns.

M.O. Why did you not go for it?
M.K. Because I thought, and still think, that I had found a good cemented solution - and cement is undoubtedly useful.

M.O. What worried you about cementless arthroplasty?
M.K. That a cementless implant would not become fixed, or that I would not be able to get it out again, should the need arise.

M.O. How do you account for the failure of implants with a high degree of canal fill and little cement?
M.K. This is a complex question. To answer it, I shall have to draw on my own experience, and to talk mechanics. In 1971, when I was using the standard Charnley, I found that implants used in small femurs, and hence only little cement, tended to be better and more durably fixed than those inserted in very wide canals and surrounded by a huge mass of cement. I also found that cement mantle fracture was never seen in DDH cases, where the canal had had to be reamed to accommodate a straight stem. In these patients, the layer of cement was, perforce, thin, and all the cancellous bone had been removed from the canal. All this argued in favour of fitting the femoral component to the dimensions of the medullary canal, which is what the Charnley-Kerboull range started in 1972 was intended to achieve. This range always came up to our expectations, and proved extraordinarily reliable in the long term: after 20 years, between 97 and 99% of the implants are still well fixed. This device also had three important features: a bright-polished surface, a rectangular cross-section, and a pronounced taper. As a result, it did not adhere to the cement; all the stresses on the cement mantle were compressive; and the rectangular cross-section applied to the entire stem prevented any torsional stress peaks. A matt finish means a microscopically irregular surface, even though it feels smooth to the touch; and this will produce a stronger bond between the cement and the implant. This, in turn, will produce excessively high shear stress levels at the cement/bone interface, to the extent that the interface may suffer fatigue failure. Also, if the implant has an oval cross-section in its metaphyseal portion, and a cylindrical one in the diaphysis, it will have poor rotational stability within the cement mantle. In fact, its rotational stability in the diaphysis will be nil, so all the torsional stress will be concentrated in the proximal third, which will load the cement to a level that it cannot cope with. I made this double mistake myself, with the CMK2 and the CMK3, partly because this design was fashionable then, and partly because the engineers were in favour. The long-term fixation of these devices has not been anything like as good as that of the MK1. I therefore think that the failure of the implants with a high degree of canal fill is not due to the fact that they espouse the dimensions of the femoral canal, but to their matt surface finish and unsuitable shape.

M.O. Do you think that one could have had problem-free thicker stems if one had kept them smooth, and given them sharp rather than rounded edges?
M.K. Yes - witness the MK3, which was kept smooth and given a rectangular cross-section. The 10-year survival rate of this cemented device is almost identical to that of the MK1. However, regardless of the dimensions of the implant, the stem must taper sufficiently for the shear stress along the length of the stem to disappear progressively and be transformed into horizontal, compressive stress. On no account should the femoral cortex be reamed in order to accommodate the largest possible prosthesis. All that is required is the removal of the cancellous bone, so as to ensure that the compressive stress transmitted by the cement goes into a rigid bony support.

M.O. This modification of the Charnley pattern has been very successful.
M.K. No doubt many orthopods had been waiting for the modified design - all those who, like myself, had found the standard Charnley ill-suited to patients with large femurs. The modified pattern was not an immediate runaway success, because the Charnley was, and still is, known as an excellent prosthesis. I also got resoundingly told off by Charnley himself.

M.O. On what grounds?
M.K. That I hadn’t understood the first thing about his device; and that if one used his implants properly, then the 2 sizes it came in were sufficient to cater for all cases. He was invited by SOFCOT, the French Orthopaedics and Traumatology Society, in 1974, and spoke for an hour and a half on how his device could be used to manage all anatomical patterns.

M.O. Did he publicly criticize you?
M.K. Not publicly, but still very severely, though not nastily. Two years later, he modified his stems.

M.O. Did he apologize?
M.K. No. Neither did he have to, because his modifications had not the same mechanical rationale as the ones I had made. He reinforced the entire stem, rounded off the corners, gave the implant a matt surface finish and two flanges, which is why it was called the Cobra. It still had a major offset and a tight neck-shaft angle. It was better fixed within the cement mantle, but put excessive stress on the cement/bone interface, which produced quite a high rate of loosening between the cement and the host bone.

M.O. He put flanges on the socket, to pressurize the cement. Why is it that you never added this feature?
M.K. Because, for a number of reasons, I do not think that it is helpful. If the flange hermetically seals the acetabulum (which it cannot do in practice), the cup cannot be pushed into the cement; and if there is the slightest leak, the pressure drops immediately. These flanges were, in fact, created for revisions with deficient acetabular bone stock. As far back as 1974, we adopted the policy of reconstructing the bony defects with bone grafts, so as to restore an acetabulum of normal size and shape. Charnley did it differently: he never used grafts, and filled the defects with cement.

M.O. What was Charnley like?
M.K. Nice, even funny when he had had a couple of drinks; very sound; enthusiastic about everything he did; and, I believe, a very rational thinker.

M.O. Did he have a sense of humour?
M.K. Certainly - except I never appreciated it, because my English was too poor.

M.O. How did your academic career advance during all those years?
M.K. The change came in 1971, when Mazas left to go to Clamart. It was about time, too, because I had been on the list of potential candidates for a Chair for three years, before the vacancy occurred.

M.O. It was Postel’s decision?
M.K. Of course, because the vacancy was in his Department.

M.O. Why should someone with your experience in hip arthroplasty still be doing trochanter osteotomies?
M.K. There are several reasons why. Firstly, I don’t like dividing or detaching muscles, because I think it weakens them and this, in turn, may give rise to instability. Secondly, this approach is very much better than others in high congenital dislocations and for revisions with bone defect reconstruction. Thirdly, if it is not well done (both the resection and the reattachment), it can produce all manner of problems. The best way to ensure that one does it well, especially if one is teaching junior surgeons, is to do it in each and every case; and if it is well done, complications will be very rare.

M.O. Twenty years ago, the argument was about femoral head diameter. Is this still a major issue?
M.K. I think it is with PE sockets, because the amount of debris produced is a function of the surface area swept by the head - so, the smaller the head, the less debris will be produced.

M.O. But with the new material combinations....
M.K. With ceramic-on-ceramic, or metal-on-metal, one can, obviously, use a larger head. The main problem with those combinations, where the head and the socket are made of the same material and where both are extremely rigid, is no longer one of friction. As I see it, it is a question of how long the very rigid socket will remain well fixed in the deformable and elastic acetabulum. I think that there is a considerable risk of a gradual mechanical mismatch.

M.O. Would an OA patient have a very elastic pelvis?
M.K. Not very elastic, I agree; but the lower part of the acetabulum would undoubtedly be elastic. That is a danger. I am not saying, though, that the danger is very great.

M.O. Have you done any cementless cups?
M.K. No, and I do not intend to do any, for two reasons. A correctly implanted PE socket that articulates with a 22.22-mm head will loosen only if the PE wears, because of osteolysis induced by the wear debris. This late loosening as a result of osteolysis around the cup will also happen with cementless PE sockets. So, I do not think that their fixation will last better than that of the cemented sockets that we are using. Also, putting a cementless socket into an acetabulum that is oval rather than round, as a result of OA, would require an enlargement of the cavity at the expense of the superior dome, which becomes a little more fragile, and above all at the expense of the anterior and posterior walls, which will be left thinner. This aggravates the mechanical mismatch between a very rigid cup and a bony socket that has been made more deformable in its lower part; and I think that this entails an appreciable risk of the cup parting company with the bone. We are now getting reports of press-fit cups that had been perfectly fixed for 8 or 9 years suddenly failing at 10 years.

M.O. What is a cemented cup like after 20 years?
M.K. A Charnley socket implanted in the 1970-75 period would most likely be intact or, at worst, very slightly worn, and perfectly fixed, at 20 years. The PE used in those days was obviously very wear-resistant, much better than the stuff that has been around over the past 15 years.

M.O. But the bone stock quality changes over the years....
M.K. Of course, but it seems that the progressive osteoporosis of old age is not enough to loosen a socket or a stem.

M.O. What did you think of the spate of French prostheses that came on the market between 1970 and 1990?
M.K. I think they show that French orthopods have a lot of imagination and creativity; unfortunately, this creativity is not always matched by an understanding of joint mechanics.

M.O. If you had to design a implant today, how would you go about it?
M.K. In 1972, I designed a prosthesis based upon Charnley’s original pattern; this device proved to be virtually perfect. In an attempt to do better yet, I have since redesigned two others, and got it wrong both times. So I am now back to the initial model, and since there is no point in going ahead with a device that has a 10-12% rate of potential loosening at 10 years, I have taken the prostheses with a matt finish and an oval or cylindrical cross-section (the CMK2 and the CMK3) and given them a bright-polished surface and a rectangular cross-section like the MK1, which remains my gold standard.

M.O. You are staying with the metal-on-PE combination?
M.K. Most of the time, yes; however, I am trying to reduce PE wear by improving the mechanical properties of the material. For a little over 2 years, I have been using a 22.22- mm zirconia head, in carefully selected cases; however, we shall have to wait at least 10 years to see whether that has been an improvement.

M.O. Take the case of a patient who is 50, and who has slight dysplasia and incipient OA. There is every hope that, with THR, he will have 20 good-quality years ahead of him. Is there still any justification for doing a shelf or an osteotomy?
M.K. CR Michel’s slogan “20 happy years with a THR” is great, but I fear that it is not always true. The non-prosthetic management of painful dysplasia in a 50-year-old is undoubtedly a difficult problem. In mixed acetabular and femoral dysplasia, a shelf plus a varus osteotomy is a sound management principle, and there is a good prospect of an excellent outcome; however, recovery from the operation is slow, and the patient will be a long time off work. This is why these patients generally dislike this option. They will then represent several years later, by which time the hip will have deteriorated to the point where THR will be required. However, there are some patients who want to hang on to their hip for as long as possible, and who can afford to take a year off work.

M.O. Do you have such patients?
M.K. Sometimes, but they are getting fewer. In any one year, I will see some 15 DDH cases who could be managed without a prosthesis, but I operate on only 2 or 3 of them.

M.O. Was it not difficult for you, in your surgical career, to forsake everything else to concentrate on hips only?
M.K. First of all, I have not been doing hips only all my working life. As a young surgeon, like many of my colleagues, I was an all-rounder. Then, as time went by, surgery of the hand, the elbow, the shoulder, the foot, and the spine, fell by the wayside, and I was virtually doing hips and knees only. It is the patients that force one into hyperspecialization. It would, of course, be boring to operate on one joint only if all I did were primary OA; however, well over 50% of my work is revision surgery, where there no two procedures are the same.

M.O. Will you accept opposite views?
M.K. I have my ideas, my convictions, which I think are well thought out and well-founded in my personal experience. I do not like unsubstantiated objections, because I think that attacking someone for the sake of attack is unproductive. However, I am definitely prepared to discuss and argue, and will go out and actively seek a debate. Did you doubt that?

M.O. Not at all. I was just wondering whether you see yourself in the juniors working in your Department?
M.K. Sometimes they are just like me. In general, though, there is an undeniable difference between the junior surgeons of 40 years ago and those working with us today. No doubt, this is because their training is different; but more probably it is due to the fact that people, and people’s minds, have changed; and that goes for the younger generation as much as for the older one. In the Department, the general climate is much more relaxed now than it was 40 years ago - and that is a very welcome change. All in all, I find that, by the end of their specialty training, our juniors are as good as we were at my time.

M.O. And the way they work?
M.K. There has been very little change. Juniors 40 years ago, like those of today, had to examine patients, keep the notes up to date, be on call, assist at surgery. In my time, juniors did less actual surgery than our present-day trainees. However, there was much less surgery being done in those days, and the juniors had plenty of time to busy themselves with the patients’ charts.

M.O. It seems that people will argue more freely nowadays.
M.K. There is no doubt that the general climate in the Department is more sociable, the seniors won’t pull rank as much as they used to, and people get on with each other better than in the past. This is, of course, a change for the better; however, this freedom must not degenerate into anarchy, because if this were to happen, the Department would cease to think and act as one.

M.O. Don’t you think that the Cochin has gradually withdrawn into its shell?
M.K. That is fair comment, especially when one thinks of the Cochin in 1960 and the Cochin now. Back in the ‘60s, there were two great centres of orthopaedic surgery in Paris - the Cochin, and Garches. Nowadays, the Cochin and Garches are but two among a dozen units, most of which trace their origins back to the two big centres, without being clones. It’s probably also due to the fact that, at the Cochin, we do not publish enough, even though we are still doing a lot of work. However, I have seen so many things that were hailed as the answer to all our problems one year, only to be forgotten again a year later, that I think it better to keep quiet unless one has something well-founded to say. This is why we don’t write a lot of papers. I think that is going to change - it will have to change, because anybody wanting to move up on the hospital career ladder needs a lot of publications to his name. Unfortunately, writing papers often means that there is less time available for learning one’s craft. This system, which values a candidate’s clinical competence and surgical skill less highly than his written output, is unrealistic and absurd. In all probability, it is going to be completely counterproductive.

(Transl KRMB)