Unicompartmental knee arthroplasty is a treatment option for patients with osteoarthritis of the medial aspect of the knee joint. Excellent clinical and functional results have been reported in the literature1,2. Patients who are managed with this procedure benefit from minimally invasive surgery, less bleeding, lower infection rates, and quicker recovery compared with total knee arthroplasty3,4. However, the Swedish Knee Arthroplasty Register showed that the revision rate is higher after unicompartmental knee arthroplasty than after total knee arthroplasty5. The most common cause for revision is mechanical loosening of one or both components6,7.
Aseptic loosening of an implant implies failure of the bone-cement or cement-implant interface. The strength of the bone-cement interface is dependent on cement penetration and interdigitation into cancellous bone8-15. However, cement penetration into bone can lead to increased interface temperatures during polymerization of the cement. Thermal damage of the interface bone may be a cause of loosening16.
Studies of hip arthroplasty with cement have shown that the longevity of an implant is dependent on a strong bone-cement interface17,18. Oxford unicompartmental knee arthroplasty with cement has had excellent long-term results in experienced hands19-21. Nevertheless, aseptic loosening of the femoral component can be a clinical problem1,2,4,7,22. In 12% to 38% of cases, isolated femoral component loosening represented the cause for revision5,7,23.
To our knowledge, no experimental studies have been published on the cementation technique for unicompartmental knee arthroplasty. The purpose of the present study was to examine the effect of pulsed lavage compared with syringe lavage on cement penetration under the femoral component of the Oxford unicompartmental knee arthroplasty, with particular consideration of the temperature measured at the bone-cement interface. Our hypothesis was that pulsed lavage improves the penetration of bone cement into cancellous bone without a risk of thermal damage to the interface bone.
The experimental study design was accepted by the local ethics committee. Minimally invasive Oxford unicompartmental knee arthroplasty (Oxford Phase III; Biomet, Swindon, United Kingdom) was performed in twenty-four paired fresh-frozen cadaver legs. All procedures were performed by one surgeon experienced with the Oxford unicompartmental knee arthroplasty surgical technique. Right and left knees were randomized into two groups with use of a computer-generated list.
After preparation of the medial femoral condyle, standardized drilling of eight anchorage holes was performed in a circle around the central peg hole to open the bone sclerosis (Figs. 1-A and 1-B).
Bone lavage was performed with use of a manual syringe (Group A) on one side and pulsed lavage (Group B) on the other side (ScandiMed Biomet, Merck, Sjöbo, Sweden) before cementing of the femoral component (see Appendix). In each group, an identical volume of 1000 mL of saline solution was used. The other steps of cementing technique were the same in both groups.
After bone lavage, the bone was dried with use of suction and then with use of two gauze pads to remove the residual fluid. The central peg hole was filled with bone cement. A layer of cement was spread over the surface of the femoral component. In both groups, bone cement (Biomet) was used after vacuum mixing (OPTIVAC M; Biomet Cementing Technologies AB, Sweden) at a mean room temperature (and standard deviation) of 20.8°C ± 0.8°C and a humidity of 61.8% ± 4.7%. The femoral component was impacted with use of a light mallet. For curing of the cement, the knee was positioned in 45° of flexion and an appropriate spacer of different thickness (feeler probe) was inserted for gentle compression.
Cement Penetration Pressure
A special subminiature pressure sensor (XPR36) (0 to 5000 kPa) with measurement amplifier (model XAM-MV; Disynet, Brüggen, Germany) was placed into the central peg of the femoral component to measure the pressure of cement penetration. Pressure data were continuously recorded during implantation of the component and polymerization of the bone cement.
Interface Temperature
A special drill guide using the spigot system was designed to place a temperature probe (length, 100 mm; diameter, 1.6 mm; analogous transmitter type, 313) (model Pt100; B+B Thermotechnik, Donaueschingen, Germany) and to measure the interface temperature exactly 5 mm from the central femoral peg hole.
Ligament Tension Force
The ligament tension force was measured with use of a force probe (1000 1 kN) (model 8413; Burster, Gernsbach, Germany), which was integrated in an original feeler gauge. The feeler gauge of the appropriate size was inserted in 45° of knee flexion after cementation and implantation of the femoral component.
Pressure, temperature, and ligament tension force data were continuously recorded in real time with use of custom-made data logging software (TEP-Force) via a USB-DAQPad (NI DAQPad-6015; National Instruments, Austin, Texas) for a total of twenty minutes after the cement mixing was started.
Evaluation of Cement Penetration
Each femoral condyle was cut into four slices of identical thickness in the sagittal plane with use of a diamond band saw (ETS 36/69 MAKRO; EXAKT Apparatebau, Norderstedt, Germany). All cut surfaces were digitally imaged under standardized conditions (EOS 350 D; Canon Deutschland, Krefeld, Germany) with constant camera settings (lens aperture, 5.6; ISO [International Organization for Standardization] 100) and focal distances (500 mm).
The areas of cement penetration on the cut surfaces were marked with a magnetic tool, and binary pictures of the cement areas were created with use of computer software (Photoshop 7.0.1; Adobe, San Jose, California). The size of the cement areas was measured in pixels with use of MATLAB (MathWorks, Natick, Massachusetts) and was converted to square millimeters (mm2) with use of a reference marker that was placed on the surface of each bone cut (Fig. 2). All measurements were performed twice by two examiners to assess the intraobserver and interobserver reliability.
Statistical analysis was performed with use of SPSS (version 15.0 for Windows; SPSS, Chicago, Illinois) and the nonparametric Wilcoxon test. The intraobserver and interobserver reliability were tested with use of the Spearman rank correlation coefficient. The tests were two-sided, and the level of significance was set at p < 0.05. All values are presented as the mean and standard deviation.
The study was planned to detect a standardized difference in cement penetration of d = 1. To achieve a statistical power of 80% and an alpha of 5%, the study would have required a sample size of eighteen per group for the nonparametric Wilcoxon test. However, because of feasibility and cost, we performed the study with use of twelve pairs of human fresh-frozen knee joints. With this modified sample size, the study was adequately powered to detect a standardized treatment effect greater than d = 1.3. The 30% difference observed in this study translates to a standardized difference of approximately d = 2.
Source of Funding
The study was financially supported by the county of Baden-Württemberg and the University of Heidelberg. Biomet supported this study with implants, cement, and surgical instruments. The funding source did not play a role in the investigation.
Aseptic loosening is the most common reason for revision of a unicondylar knee prosthesis24 and accounts for about 50% of such revisions7,25. Loosening is the second most common cause of failure following Oxford unicompartmental knee arthroplasty with cement26. Failures of the femoral component are the result of poor initial fixation25. In clinical practice, loosening of the femoral component is very difficult to diagnose as the interface cannot be seen on radiographs because it is hidden by the femoral component22,23. Loosening of a cemented implant indicates that the bone-cement or cement-implant interface has failed27. Primary stability is regarded as an essential requirement for increasing the strength of the interface28 and therefore the longevity of a cemented prosthesis29. Thin cement mantles and cement mantle defects are associated with increased loosening of the components30,31. Several studies have shown that implant stability improves with deeper cement penetration into the cancellous bone8,9,11,28,32-35. Cleansing this bone bed prior to cementation can improve cement penetration34,36,37. Krause et al. showed that the lavage depth limited the depth of cement penetration in human tibiae9. Breusch et al. showed that the use of jet lavage yielded significantly improved cement penetration compared with conventional lavage in a cadaveric study of cemented hip stems13.
To our knowledge, there have been no experimental studies showing the effectiveness of pulsed lavage at the time of unicompartmental knee arthroplasty involving femoral component fixation with cement. Our results demonstrate that cleansing the cancellous femoral bone bed with use of pulsed lavage is far more effective than syringe lavage and leads to deeper cement penetration. The cement penetration areas that we found in the bone cuts were significantly larger in the pulsed lavage group. Cement penetration was improved by approximately 30% in comparison with that following syringe lavage. However, cement penetration is dependent on pressurization8,38. Direct pressurization of cement, which is a mandatory part of modern hip cementing technique, is not possible in unicompartmental knee arthroplasty because of the more complex anatomy of the knee and the minimally invasive approach. Pressure is usually applied indirectly in three phases: (1) during the filling of the central peg hole, (2) during insertion and impaction of the femoral component, and (3) as a result of the tension of the ligaments when the feeler gauge is inserted with the knee in 45° of flexion. Phase 1 was standardized in all cases, and our measurements were able to record Phases 2 and 3 in real time. We found similar cement penetration pressures and ligament tension forces in both groups.
We cannot state the optimum penetration depth of bone cement for unicompartmental knee arthroplasty. Ebramzadeh et al., on the basis of radiographic analysis, concluded that 2 to 5 mm is the optimum depth for total hip arthroplasty with cement18. Joshi et al. observed a decreased rate of local osteolysis at penetration depths of at least 3 mm39. Askew et al. considered 4 mm of cement penetration to be optimal10. A deep cement penetration and a homogenous cement mantle under the femoral component are probably best for a permanent transfer of the applied load and shear forces into bone.
Increased cement penetration, however, may be associated with thermal damage of the interfacial bone because heat is generated during the process of polymerization of the bone cement. Oates et al. and Sih et al. feared that heat necrosis of the bone bed would occur if cement penetration was ≥3 mm, but they also concluded from their studies that the benefit of higher cement penetration depths is greater than the risk of tissue damage40,41. There have been several studies in which no evidence for heat necrosis caused by bone cement was found42-44.
In our study, the interface temperatures measured in the femoral bone were significantly higher in association with the use of pulsed lavage compared with syringe lavage (p = 0.028). However, our data clearly showed that with a maximum interface temperature of 25.7°C, these temperatures never reached critical values of 44°C. Eriksson and Albrektsson stated that the regeneration of bone is affected if temperatures of at least 44°C to 47°C persist for more than one minute45,46. The interface temperatures measured in our series are very unlikely to cause necrosis of the interface bone. However, the maximum interface temperature that was measured seems low in comparison with the body temperature of 37°C. Body temperature may well alter the result. Preheating of the cadaver knees would have eliminated this argument, but we measured the bone interface temperatures during Oxford unicompartmental knee arthroplasty procedures in the operating room with use of an infrared thermometer (Inspacto 900 plus; Infrapoint, Saalfeld, Germany) and obtained temperatures well below the body temperature, between 22.00°C and 26.90°C, probably because of exposure of the bone surface during the procedure.
Our study has some limitations. Bone cement penetration depends on more factors than just the type of bone lavage. By using fresh-frozen cadaver knees, we tried to simulate a realistic perioperative situation, but bleeding could not be simulated. Bleeding may have altered the result, but as Oxford unicompartmental knee arthroplasties are performed with the hip flexed to about 30° and additional use of a tourniquet, bleeding would appear to be a negligible aspect. The structure and density of the bone as well as the preparation of the cancellous bone, the cement application and timing, the pressure applied during cementation and curing, the viscosity of the bone cement, cement mixing, temperature, and humidity are all important factors for the result of the cementation. We controlled and standardized these factors in our experimental setup.
In conclusion, we recommend using a thorough pulsed lavage of the femoral bone bed as a routine procedure during unicompartmental knee arthroplasty with cement to improve cement penetration without the danger of thermal bone necrosis.
Disclosure: One or more of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of an aspect of this work. In addition, one or more of the authors, or his or her institution, has had a financial relationship, in the thirty-six months prior to submission of this work, with an entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. No author has had any other relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.