In his latest Clinical Update, Gary Culliton examines suggestions that there is limited evidence for orthopaedic knee implants and apparent impediments to clinical trials.
Call for ‘bravery to publish’ after failures
Knee implant companies are selling a product that may have limited data to support its survival, a leading Dublin orthopaedic surgeon has suggested. “However, side-effects have emerged in regard to drugs too — after they have gone through Phase 1, 2 and 3 trials,” said Hermitage Clinic surgeon Mr Fergal McGoldrick.
“A suggestion that 10,000 guinea pig patients might be selected for five years and that on that basis, a decision would be taken that the product is safe for everyone to use, lacks insight,” he commented.
Mr McGoldrick added: “However, people — particularly family doctors — need to be told if there is a high morbidity in regard to knee replacements or if there are failures. If patients are not told the outcomes and only discover them when a high failure rate is reported in the media, that’s wrong.”
Replacements are up
Epidemiologically, knee replacements are going up but that is not because there is a dramatic increase in the diagnosis of arthritis — outcomes for knee replacements have been shown to be successful.
Nevertheless, doctors are evaluating outcomes in orthopaedics far more than in any other specialty, Mr McGoldrick maintained: “I don’t know of any procedure by any doctor that has comparative data,” he said.
Ireland follows the UK model: this country uses the same implants and applies the same standards. “There are one million patients on the Joint Registry in England and Wales at present,” said Mr McGoldrick.
“It is the biggest registry in the world and includes outcome measures from orthopaedic surgeons. The Registry mentions individual prostheses and seven-year survival data and failure rates for individual implants are shown.”
To suggest that orthopaedic implants are put onto the open market without a good deal of research is not true, the surgeon said, and there were reasons why traditional clinical trials could not be conducted on prostheses, he insisted.
Implants are tested in the factory through failure: “Anatomically, we’re stuck. The nearest thing to a human hip is a pig’s hip. There are no bipedal animal models for comparison. There is no animal knee that can be compared to the human. However, in the modern era, a conceptual design can be bio-engineered; it can be fabricated and repetitive stress-loading tests (until failure) are done,” said Mr McGoldrick.
“It is difficult situation,” he said. “How do you evaluate evidence-based outcome measures on human beings? There is no model for that. If you do not have a comparable model, how do you do it? We don’t have a humanoid monkey that we could do knee replacements on for five years, to see whether they work before we put them into humans. You cannot do the Phase 1, 2 and 3 equivalents of a drug trial. How many subjects might you pick and for how long might you study before you decide it is safe to use an individual new prosthesis? That’s the limitation. Instead, you design a knee or hip, do all the bioengineering studies on it and then you approach the surgeons.”
A trial of the prosthesis on 100 patients might ensue at a particular surgical centre and new prostheses are usually tested using multi-centred trials, involving surgeons who do high numbers of procedures.
“My concern is whether the individuals in multi-centre studies receive inducements,” Mr McGoldrick said.
“Do the surgeons have a vested interest? The independence of people presenting papers has to be questioned. Currently, surgeons — usually affiliated with a university — will trial a prosthesis themselves and begin doing some form of outcome measure. They will do the study longitudinally, over four-to-five years, and publish. Journals are unlikely to be interested in data covering less than three-to-five years, when the data would have some scientific merit. If the surgeon finds the data deviate significantly from what is expected, the data may be published much earlier than planned. What matters is surgeons being brave enough to publish their outcomes where they are having failures. It is only when this happens that an alert goes out that there is a problem with the prosthesis,” explained Mr McGoldrick.
Outcomes have to be published at six months, a year, two years and five years and most orthopaedic surgeons nowadays are highly skeptical in regard to outcomes and what they read in the literature, Mr McGoldrick stated.
“Primarily nowadays, surgeons want to know the outcome measures from the Joint Registry,” he added. Mr McGoldrick said surgeons probably have to indicate their independence on this issue — that they have no financial vested interest. Generally though, they are not tied-in financially, he said.
Most doctors will concede that in cases of inflammatory synovitis — as distinct from degenerative osteoarthritis — steroid injections in a patient will work and will reduce the inflammatory symptoms. However, they will not cure the problem and steroids will not work if there is established osteoarthritis. Where there is inflammatory tissue — synovitis — use of non-steroidal medication, exercise, physiotherapy and weight control are helpful and anti-inflammatories can be tried. An off-loading caliper that tilts the knee off the bone can also be useful.
“Where there is degenerative arthritis — a worn joint (bone on bone) — none of those modalities can work,” Mr McGoldrick added.
“In 40- to 55-year-olds, where there is bone-on-bone contact confirmed by arthroscopy, there is not a choice. At the end of the day, the joints have to be resurfaced. Arthroscopic keyhole surgery can work in patients who have transitional degenerative wear.”
Kellgren grading system
The Kellgren grading system looks at the surface wear and it gauges the depth of articular cartilage loss. In early degenerative changes, keyhole surgery can be beneficial and when advising patients of the advantages of partial or minimally-invasive surgery, it should be pointed out that the outcome for a partial is inferior to a total knee replacement, said Mr McGoldrick.
There is a discrepancy in outcome measures of 6 per cent between partial and total knee replacements. British Joint Registry measurements for survival of partial knee replacements (92 per cent) are inferior to total knee replacements (98 per cent) at 10 years and 6 per cent of patients might have avoided a revision, had they had a total knee replacement.
Score being tested at 27 Irish centres
The incidence of degenerative arthritis in the knee has always been far greater than in the hip. It was only when the Hospital for Special Surgery in the US came up with a knee replacement, invented by John Insall in 1974, that the modern form of knee replacement emerged.
Early experience with knee replacement did not follow hip replacements (which took off in the 1960s and were initially successful). The first and second generation of knee arthroplasties in the early 1970s failed. Most loading studies on prostheses involve repetitive strain analysis (RSA) studies.
The objective is to calculate how many stress cycles are required before an implant — be it hip or knee — fails. Four generations of plastics have been used over the years and ultra-high molecular weight plastics are now used.
The orthopaedic community internationally, led by the Swedes, has decided to monitor for itself to see whether implants work.
A total of 27 centres in Ireland are being tested to see how accurate the Oxford hip and knee score is, prior to the introduction of the hip and knee registry here.
The registry in England has been in operation for nine years, while the registry in Sweden, by contrast, goes back to 1979. America has no hip or knee registry and that country can be criticised for a lack of standardisation, in the view of Hermitage Clinic surgeon Mr Fergal McGoldrick. However, there is constant surveillance and recalls have been ordered by the Food and Drugs Administration (FDA).
The FDA has sanctioned joint replacement manufacturers in cases where there were failed implants. In cases where there are significant recalls of faulty implants, the manufacturers were penalised and fined.