Knee replacements have been implanted for over 50 years and in serious numbers since the 70s. The fundamentals of metal and plastic construction have not changed, nor will they for the foreseeable future. Advances in stem cell therapy and similar technologies have a long way to go to provide structural articular support in the knee. So they really only have a role in reducing inflammation and regeneration of cartilage or neo-cartilage in isolated  articular defects.

Knee replacement despite all our advances and refinements has a high dissatisfaction rate. The Swedish Knee registry published its results on 27000 TKAs, in 2000 and had a dissatisfaction rate at 2-17yrs  of  8%. A paper by Robertsson et al. in 2003 found similar results but only 4% for THA.

The recent literature has described dissatisfaction rates at up to 15%.

Why have the results deteriorated?

Over the same period Survivorship has gone up from the high 80%s at 10 years to 92% at 14 years in the current AOA NJRR.

The current TKAs on the market all have greater than ten year data and most of the poor performing TKAs have been withdrawn from the market. So despite the improvement in prostheses and survivorship patients remain unhappy.

Over the last ten years perioperative pain management has improved dramatically, and the use of Tranexamic acid to reduce bleeding postoperatively has dramatically reduced the need for transfusion, and most probably has contributed to reduced bruising, swelling and pain. We have intensive rehabilitation programs with physiotherapy and hydrotherapy and improved support networks. We have instituted comprehensive patient care models with education and nurse case managers. Nobody goes to theatre being under any illusion as to how tough the recovery will be after TKA.

So how do we explain this anomaly?

Certainly we are operating on a different patient demographic than 20 years ago. We have younger more high demand individuals and older patients with more co-morbidities. Expectations are high of both our patients and ourselves.

Also the tools used to assess satisfaction have become more sensitive, and perhaps these factors explain some of the increase in dissatisfaction.

We know greater age, more co-morbidities and more pain prior to surgery are predictors of poor outcomes. Psychological profiling has revealed particular personality types whom will have a poor outcome, in particular those with depressive illness.

Also, despite our advances in design, this still remains metal on plastic with limited size matching and poor compensation for removal of ligamentous structures. This is obvious when you compare the dissatisfaction of TKA and THA, which is roughly double. Stability and ROM is easier to consistently achieve in a ball and socket joint.

Where do we go from here?

If we can identify patients whom are going to, or are highly likely to ,have a poor result should they be offered a TKA? This is a topic for the Health Economists and ultimately society as a whole to grapple with in years to come. At this stage patients need to have risk factors identified and be counselled accordingly, about the reasonable expectations they should have after surgery.

There are clearly advances to be made in design, but this will be evolutionary rather than through new design philosophies. The Regulatory network and litiginous state of society make it extremely difficult to innovate in Orthopaedics at this moment. Also R%D costs add to the cost of implants which is a hot topic considering the potential numbers to be implanted with our ageing population.

Our understanding of alignment issues and instability are improving, and so is the technology, ie. Computer navigation and possible Patient Specific Implants [although the later remains marketing hype rather than reality, as we are still using the same 7-8 sizes for each component.]  to achieve improvement in these parameters.

There are certainly ongoing challenges but despite all the shortcomings, there are very few patients whom report the result is worse than prior to surgery. Even the Health Economists recognise the benefits, when measuring the relative benefit of this intervention against others with QALYS [Quality Adjusted Life Years].

Introduction

LLD after THA remains extremely common and is one of the leading causes of dissatisfaction. It has been etched into orthopaedic folklore that LLD is an inherent risk of THA.

On Feb. 27, 2009, the AAOS, Committee on Professionalism (COP) Hearing Panel conducted a hearing;

The panel censured a surgeon whom provided an opinion regarding standard of care and leg length discrepancy against a colleague. However he was condemned because they felt the performance fell within generally accepted practice standards as no established standard for leg length discrepancy existed.

I think this bears closer scrutiny. The excuses for LLD are the lack of reliable intraoperative methods of measurement, complex anatomical variation, deficiencies in prosthesis inventory and the requirement to ensure stability over LL.

Over time these arguments are becoming suspect.

There is no doubt the incidence is getting less and the threshold of unacceptable with it.  Prosthesis inventories allow a wide range of length and offset in femoral stems and larger heads mitigate the risk of instability. There are accurate templating systems for standardized xrays and CT Scans can forestall any unpleasant surprises with dysplastic joints. There are many leg length devices on the market and intraoperative xrays are available in the majority of medical facilities.

What is Limb length discrepancy or inequality [LLD]

Definition: A condition in which paired limbs are noticeably unequal. In Lower limb it is referred to as leg length discrepancy.

Natural Incidence: Varies between 40% according to Subotnick in 1981 to 70%  by Woerman in 1984.

LLD > 20mm affects at least 1 in 1000 people according to Guichet  in 1991

A meta-analysis by Knutson in 2005 found an average 5.2mm with 90% of the population having a LLD.

Classification:

Structural: Defined as those associated with shortening of the bones. These include Congenital causes, Infections, Paralyses, Tumours, slipped femoral epiphyses and Total Hip Replacement.

Functional: Defined as those that are the result of altered mechanics of the lower extremities. These include Foot malalignment, knee hyperextension, hip abduction tightness or weakness.

How do we Measure LLD?

The most reliable method, to within 1mm, is CT scanogram., according to Aaron in 1992.

Alternatively, Functional assessment involves standing the patient on blocks and building up the length of the short leg.

What constitutes a significant discrepancy?

There is disagreement as to the absolute amount that warrants treatment, with authors suggesting greater than 30mm and some as low as 20mm. These figures are mostly derived from the paediatric literature.

What we do know is that the more active a patient the smaller the discrepancy that can lead to symptoms. In a normal population 10-25mm rarely causes symptoms according to Siffert in 1987, but Friberg in 1982 found an increased incidence of stress fractures in Finnish Army conscipts with as little as 10mm. Subotnick in his review suggested ¼ inch is tolerated by athletes compared to ¾ inch in the normal population.

There has been much written about gait and the altered joint reaction forces, which can lead to increased energy consumption.

However, functional limitations with standing posture and balance, impact on walking and running have all been reported but there is considerable disagreement and it is difficult to draw conclusions.

Leg length discrepancy and Low Back Pain

There are several papers supporting the premise that LLD contributes to LBP, but statistically it was only valid if >20mm.

Several studies have linked LLD to Scoliosis but one author thought it was less clear.

Leg length discrepancy and hip pain

There is extensive literature supporting an increase in hip pain in LLD.

Conclusion

The younger and more sedentary the person the greater the LLD discrepancy tolerated, whereas the converse is true with active individuals with those having a later onset experiencing greater symptoms, even with LLD <20mm.

Incidence of LLD after Total Hip replacement

This is a common outcome and reported as usually <10mm.

Literature pre 2002

Bruce Love in 1983 reported that 18% had greater than15mm of LLD and 6%  of the patients required a shoe correction.

Wylde, retrospectively, reviewed 1114 patients from 1993 to 1996 and he found 30% complained of LLD and 49% of these patients were bothered by the difference.

Austin’s data in 2000 was indicative of a changing trend and revealed an Average LLD after THA ranged from 2.8mm to 11.6mm but many authors still reported larger ranges and occasional outliers.

Literature post 2002

Showed a progressive improvement with increasing focus on Leg Length and the use of  leg length measurement devices, sophisticated templating software and intraoperative xrays.

Garcia-Juarez in a review article in 2013 and noted that usually <5% of THRs had a LLD greater than 20mm in contemporary literature.

Functional disturbance 

Roder in the Journal, BMC Musculoskeletal Disorders 2012, collected data from 15 centres in Switzerland on 10415 THA, and found 405 with a 10mm LLD and 73 with greater than 10mm LLD. There was a 4% incidence of Leg Lengthening in this group. This was shown to be statistically associated with reduced walking capacity, limp and decreased satisfaction. It did not, however, correlate with pain.

Edeen  noted a LLD of >9.7mm in 68 patients. 32% of these patients were aware of the discrepancy and disturbed by this difference.

Beard DJ as part of the EPOS group, used the Oxford Hip Score, in 987 THA, and found a significantly worse functional result at 3 years post THA if LL was >10mm.

Weng WJ produced the same findings. Apart from the Oxford scores being worse there was no improvement over time.

However, not all authors shared this view, White and Dougall showed no correlation between LLD and functional outcome or patient satisfaction.

The leg does not necessarily need to lengthened, some patients can be disturbed by the leg being too short on the operated side.

Interestingly Krestic found that some patients even without a measurable LLD perceive a LLD.

Real and Apparent LLD

It is important to differentiate between real and apparent LLD, as the later is often related to a lumbar scoliosis and can be improved over time with physical therapy. Real LLD is due to component mal-positioning.

Conclusion

The available literature shows a decreasing tolerance, in the Orthopaedic community, to LLD over time as it has become recognized that LL >10mm results in functional disability and poor satisfaction.

There appears to be little argument that >20mm is most frequently associated with significant functional disturbance. It is important to note that even patients with no measurable LLD can perceive a LLD

Methods of preventing LLD

“There has always been an argument that stability takes precedence over length and compromises may have to be made in the later to achieve the former, but with the flexibility in modern designs and careful planning this should rarely be necessary in primary THA.” [ Maloney WJ et al., J of Arth 2004]

Pre-operative clinical assessment of LLD and assessment of spinal disease and pelvic obliquity are essential parts of the pre-operative assessment. Significant pre-operative contractures or deformity may not resolve and result in an apparent or functional LLD, despite the real LL being re-established at the time of surgery. Recognition of this possibility and appropriate counseling of the patient is important in this circumstance.

Accurate radiological assessment of LLD requires standardized Xrays and sometimes standing films.

Accurate pre-operative templating, either against the operated hip or contralateral hip if there is too much deformity, is essential. The goals are to re-establish the hip centre of rotation, the level of neck resection, neck angle and neck length.

Intraoperative assessments include measuring the Acetabular cup placement against the true acetabular floor and transvers acetabular ligament, measuring the resected neck against the prosthetic neck trial component, but the later does not prevent stem being placed high in the neck and subsequently being longer.

Intra-operative assessment of length using the relative position of knees and feet when in a lateral position is seldom accurate because of the preoperative adduction and shortening of the operated hip, which if corrected will inevitably result in the leg being longer. In the supine position it is easier to stabilize the pelvis and assess alignment but it is still difficult to accurately assess the length through the drapes. Other intraoperative techniques have been reported using different measurement points with variable accuracy  [Kersic M et al., J of Arth. 2014].

Pins and Calipers are widely used and can provide accurate leg length and femoral offset, but pin placement can be dislodged or moved intraoperatively.

The advent of minimal invasive or muscle sparring approaches has led to greater difficulty as the normal anatomical landmarks are obscured and the potential for LLD is greater.

Intra-operative assessment using Image Intensification or single shot plain X-Ray are increasingly used, and most surgeons request a postoperative Xray in recovery.

While all these techniques have minor limitations a combination of those techniques, mentioned above has a high probability of avoiding LLD beyond 5mm. [Hofmann AA et al., Am J of Orthop. 2008]

The use of Computer navigation  or Patient Specific Implants are emerging technologies for hip replacement have yet to achieve a level of accuracy or availability for wide use amongst Orthopaedic Surgeons.