Healio
Healio
October 01, 2008
13 min read
Save

One-year Follow-up of 214 Total Knee Arthroplasties With Navigated Columbus Implants

Abstract

In this study, 206 patients with 214 Columbus total knee arthroplasty (TKA) implants were followed up at 1 year. Preoperatively, patient demographics (mean [SD]) were 85 male; age, 69.7 (8.7) years; ASA score, 2.5 (0.7); body mass index, 32.2 (5.7); 161º varus and 27º valgus; fixed flexion, 5.6° (6.1°); flexion, 96.1° (18.8°); and Oxford score, 43 (7.0). At 1-year follow-up, results were fixed flexion, 0.9° (2.6): maximum, 17°, minimum, 0°; flexion, 101.3° (9.1): maximum, 125°, minimum, 75°; and Oxford score, 23 (7.7). Radiographs showed radiolucent lines in 6 femurs in 1 zone; 1 in 2 zones and 0 in more than 2 zones; and 3 tibias in 1 zone. There were 2 deep infections. Ninety-eight percent of patients were satisfied with their TKA.

More than 65,000 total knee arthroplasty (TKA) operations are performed in England, Wales, and Scotland each year. Despite the overall success of this procedure, the outcome depends on several factors including patient selection, implant features, surgical technique, and postoperative follow-up. In 2003, the Agency for Healthcare Research and Quality (AHRQ) published a comprehensive report 2 that concluded that there was a lack of experimental methodology and longer-term results to categorically deduce TKA outcomes.3 This series reviewed 214 computer-assisted TKAs, all carried out in a dedicated arthroplasty center, after 1 year. Notwithstanding a short-term result, we believe that we have an extremely strong experimental methodology. From preassessment through follow-up, this is a review of systematically collected data of all patients seen in the arthroplasty center. All patients received the same standard fixed-bearing knee design (Columbus; B. Braun Aesculap, Tuttlingen, Germany) implanted with the same image-free navigation system (OrthoPilot; B. Braun Aesculap) under similar anesthetic protocols. More importantly, the Arthroplasty Service at the center ensures high percentage, consistent quality follow-up that is independent from the surgeon. This has allowed us to assess the incidence of the many short-term problems associated with TKA (infection, deep vein thrombosis, anterior pain) with this particular implant and surgical set-up. This is the first study reporting 1-year review of a series of Columbus-navigated TKAs. We compared our results with current series with similar condylar implants.

Although knees have been navigated for over 10 years, navigation is not yet a routine practice. Concerns about technical difficulties and the potential increase in complications have been raised in some reports, with the investigators concluding that navigation was either a disruptive technique or too new to be used on a regular base.24,25 We hypothesized that using a nonimage-based navigation system did not affect our practice, the number of complications, or outcomes.

Materials and Methods

A prospective continuous series of 206 patients underwent 214 Columbus- navigated TKA from March 8, 2005, to December 17, 2006, after informed consent was obtained from all patients. This included all patients assessed as suitable for TKA. Of these TKAs, 93% were performed by one surgeon. Preassessment selection was rigorously performed by an arthroplasty team including physiotherapists, occupational therapists, arthroplasty nurses, senior hospital officers, anesthetists, and consulting surgeons. During the preassessment, extensive patient demographic data (Tables 1, 2) and medical history including preoperative function were collected. Preoperative radiographs including weight-bearing hip-knee-ankle, lateral, and skyline view were taken.

Table 1: Demographics of Age, BMI, and Scores for Patient Cohort

Table 2: Demographics of Gender, Side, and Etiology for Patient Cohort

Of patients in this series, 60% were women. The average ASA score (2.5) indicated that most patients were very ill/fairly ill (ASA: 1 = 12; 2 = 92; 3 = 92; 4 = 10; total 206). The majority of patients (87%) had end-stage osteoarthritis, and there were more varus knees (83%). Seventy-nine percent had an Ahlback score (1-5) ≥4, and 99% had a Kellgren-Lawrence score (1-4) ≥3.

The usual anesthetic protocol included an epidural with 1.5 g of cefuroxime prior to skin incision. The patient was placed in the supine position, and the tourniquet was inflated at 300 mm Hg for the duration of the operation, when there were no neurovascular contraindications. The knee was flexed to approximately 90° and stabilized on the side with a lateral support and a sandbag underneath the foot. A midline skin incision was made, followed by either the medial parapatellar approach in most of the knees or by the lateral parapatellar approach in fixed valgus contracture knees.

The OrthoPilot navigation system was placed at the opposite side of the knee 2 m from the knee, and the LED active trackers were affixed solidly into the femur and the tibia. After registration that included kinematics and anatomic landmarks acquisition, the knee was assessed, initially in optimal extension and then in varus and valgus stress. The knee was then assessed from the maximum extension to the maximum flexion. Before making bony cuts, a release was performed according to the classification given by Unitt26 and the measured preimplant varus/valgus stress results.12 This gave a knee that was adequately balanced within ±3°. The flexion/extension angle was recorded and the tibial cut made. It was usual to cut 10 mm on the lateral side with respect to the normal joint line in varus knees and 8 mm in severe valgus knees. The normal joint level was recorded using the pointer. A plate probe equipped with a tracker allowed the checking of the actual bone cut resection and the measurement stored. The distal femoral cut was then performed using the computer-guided jig. Initially, 9 mm was cut from the distal femur, which was the distal thickness of our prosthesis. The plate probe was then placed flat against the actual cut and the reading stored. An additional 2-mm or occasionally a 4-mm distal femoral resection cut was made in case of non-correctable flexion contracture after adapted release. Rotation was usually set according to the Whiteside line but controlled with the computer.

Table 3: Comparison of Pre- and Postoperative Functional Measures

The trial implants were set, and the assessment was performed under navigation. The leg was placed in full extension, and the varus/valgus stress measured. The range of motion of the knee from full extension to maximum flexion were assessed. Once satisfied with full assessment using the trial implants, the procedure was completed by cementing both femoral and tibial components using antibiotic-impregnated cement. Most implants were cruciate-retaining Columbus implants. A deep dish tibial insert rather than a standard inlay was sometimes used when the posterior cruciate ligament (PCL) seemed weak. Rarely, when there was still significant flexion contracture or a completely missing PCL, a posterior-stabilized implant was used. Seven patellar releases were performed, and none were resurfaced.

At the end of the procedure, with actual implants, the last assessment was recorded with knee in full extension, varus/valgus stress, and all the way from full extension to maximum flexion.

The wound was closed with either clips or sutures in three layers and then covered with a hydrocolloid dry dressing (DuoDerm; ConvaTec, Flintshire, United Kingdom) and a hydrogel inner layer (Aquacel, ConvaTec) without any drainage. Patients were mobilized on the day of surgery or the day after, under a physiotherapist’s supervision, using a Zimmer frame and then crushes or walking sticks. The postoperative protocol was standardized including early mobilization with full weight bearing and rapid range of motion (ROM) recovery. For thromboprophylaxis, the majority of patients received impulse boots, stockings, and aspirin 150 mg for 6 weeks but some had additional chemical thromboprophylaxis. Patients were discharged when occupational therapists and physiotherapists considered them autonomous and safe to return home. Outpatient physiotherapy was prescribed if needed, mainly in cases of unsatisfactory ROM. The dressing was changed before discharge and at the patient health center. Infection control nurses and arthroplasty nurses called the patients within the first 15 days to monitor progress. The wound was also assessed at the arthroplasty follow-up at 6 weeks.

Figure 1: Distribution of pre- and postimplant coronal plane alignment angles as measured by the navigation system
Figure 1: Distribution of pre- and postimplant coronal plane alignment angles as measured by the navigation system. Wilcoxon test P < .001 between groups (n = 110).		 Figure 2: Distribution of pre- and postimplant sagittal plane alignment angles as measured by the navigation system
Figure 2: Distribution of pre- and postimplant sagittal plane alignment angles as measured by the navigation system. Wilcoxon test P < .001 between groups (n = 108).

Extended scope practitioners, who have undergone extensive training, staff our Arthroplasty Service. They organize follow-up for all patients at 6 weeks and 1 year. The follow-up assessment was carried out by independent practitioners who asked patients to complete an Oxford score questionnaire to assess ROM, the wound, general progress, and satisfaction and to identify any postoperative complications. Patients were referred to consultants for review only if a problem occurred. All data were then recorded in our proprietary electronic audit system, which is accessible only to the arthroplasty practitioners. An independent consultant reviewed radiographs randomly and measured the coronal mechanical femoral axis, sagittal mechanical femoral axis, coronal mechanical tibial axis, sagittal mechanical tibial axis, coronal femorotibial angles, and radiolucencies.

Statistical analysis was carried out using SPSS version 15.0 (SPSS Inc, Chicago, Illinois). Pre- and post operative data were compared using a Wilcoxon test.

Results

Of 206 patients, 203 were reviewed (211 knees). Two patients were lost to follow-up and 1 patient died during the year from causes unrelated to the knee surgery. After 1 year, 98% of the patients were very satisfied or satisfied. Of the 4 patients who remained unsure or unsatisfied, 1 was diagnosed with Alzheimer’s disease and needed a difficult two-stage revision prosthesis for deep infection; one had a 17º flexion contracture; one with anterior pain underwent computed-tomography (CT) investigation; and one experienced problems with the other knee (loose unicompartmental knee that has since been revised).

No correlation was found between age, gender, BMI, degenerative stages, and satisfaction outcome.

The pre- and postoperative (1 year) Oxford scores, ROM, and fixed flexion from the clinic (Table 3) and the pre- and postimplant mechanical femorotibial angle and fixed flexion from the navigation system (Figures 1, 2) were compared. Wilcoxon test P < .001 between groups (n = 108).

Table 4: Postoperative Complications

Few complications occurred (Table 4 and listed below). Two severe deep infections required revision.

Average postoperative Haemoglobin was 11.1g/dL (1.5) and 6 of our patients received blood transfusions.

Discussion

This is a 1-year review of a prospective continuous study of 214 Columbus-navigated TKAs. Two patients were lost to follow-up, and one patient died of causes unrelated to the knee surgery. When asked at follow-up, 98% of patients were either satisfied or very satisfied. Among the four unsure or unsatisfied patients, three were functionally unsatisfied. It is known that the functional score continues to constantly improve after a TKA up to 2 years even though the literature cannot support specific recommendation about which patients are most likely to benefit from TKA and what type of implant or implant fixation method is most beneficial.3 Our series showed statistically improved Oxford scores, confirming results from previous studies.1,4 This study also confirmed that age, sex, BMI, and degenerative stage cannot predict a satisfactory outcome.3

Postoperatively, nine patients (4.2%) had an intermediate flexion contracture of between 6° and 19° according to Ritter’s criteria.6 In his review of 5622 TKAs, Ritter6 found a preoperative flexion contracture rate of 26.5%. According to McPherson5, flexion contracture after 1 year can continue to improve. McPherson5 showed positive outcomes of more than 10º up to 3 years after surgery. This means that even the most severe cases (17°) can improve. Average flexion was 101.3° postoperatively, with ROM significantly improved. These figures are similar to those published by Walker, who reported average ranges of flexion angle from 100° to 110° after condylar knee surgery.7 Some of the disappointing flexion angles were related to incomplete removal of posterior osteophytes (Figure 3) or limited preoperative flexion.

Figure 3: Radiolucencies measured on the postoperative radiographs for femoral and tibial components
Figure 3: Radiolucencies measured on the postoperative radiographs for femoral and tibial components. Yellow numbers show counts of radiographs with radio-lucencies in that zone.

The primary postoperative development after TKA is femoropatellar complication, which ranges from 1.5% to 12% in incidence according to Lombardi.8 In this series, two patients reported anterior pain (<1%), one underwent a secondary patellar resurfacing, and one underwent CT investigation to assess femoral and tibia rotation. The use of computer navigation may explain this low rate of anterior pain. In 2004, Stockl9 and Chauhan10 showed that computer navigation improved the femoral rotation position, and in 2006, in an experimental setting comparing various femoral rotation landmarks, Siston16 showed that even when using navigation, inconsistent femoral rotation can occur. However, he also showed that Whiteside’s line still could be a reliable reference especially with use of computer guidance. We demonstrated a similar outcome in 2007.12 Computer navigation may improve alignment and may also decrease morbidity and complications such as operative bleeding.21 Only 2.4% of our patients underwent transfusion, which is less than that reported in recent series such Sundaram’s,20 claiming 8% transfusion in a group of 200 patients.

Table 5: Postoperative Radiograph Alignment in Coronal and Sagittal Planes

In a study by Callahan,1 the mortality rate was 1.5% and the revision rate was 3.8%. In our series, two implants were revised for deep infection, one for patella resurfacing, and none for loosening, resulting in a 1.4% revision rate. In a 2008 article by Katz et al13, that compared high to low TKA volume hospitals, he confirmed his previous conclusions12 that high volume and special units are associated with fewer complications. Katz performs more than 200 TKR procedures yearly, which corroborates the following results (our results are shown in parentheses). The complication rate for mortality after TKA was between 0.53% and 0.62%.14,15 After 1 year, one patient died from causes unrelated to surgery (0.4%). The pulmonary embolus rate was between 0.41% and 0.77%.13,15 One patient experienced a nonfatal pulmonary embolism (0.4%). Deep infection rates are usually between 1% and 2%.27 Two patients (≤1% of the series) had a deep infection and underwent two-stage revision (both had BMI ≥35 and Charlson comorbidity index ≥2, which significantly increased the risk of infection). Infection remains the most challenging risk factor in TKA. Our patients are reviewed by their general practitioners within the first 2 weeks of surgery. Seven patients had antibiotics started by their GPs because of red wounds and other infections (chest, urinary tract infection). All these patients were followed clinically and had blood test at 6 weeks that ruled out TKA infections. In our series, the overall complication rate was 4% (compared with 18.1% in the Callahan review and 5.4% in the most recent review.2,3 The main drawback of the navigation was operating time. The average operating time was 80 minutes, which is similar to the published results for navigated TKR.11 Navigation slightly increases the standard operating time by 10 to 20 minutes because of the tracking setting and registration.

Figure 4: Grouping of alignment angles measured within 2° on postoperative radiograph for measurement of the coronal mechanical femoral axis
Figure 4: Grouping of alignment angles measured within 2° on postoperative radiograph for measurement of the coronal mechanical femoral axis, sagittal mechanical femoral axis, coronal mechanical tibial axis, sagittal mechanical tibial axis, and coronal femoro tibial angle showing how many patients had good alignment in all planes (n = 206).

The main advantage was to reproducibly align the TKAs. Postimplant measurements taken at the end of the surgical procedure using computer measurements showed excellent alignment in both the sagittal plane (1.4° ± 2.2°) and the coronal plane (0.1° ± 1.4°). These results were confirmed at the 6-week, long leg film review (Table 5). Using Jenny’s criteria22 to quantify the alignment reproducibility, we found that 71% of our implants had four component angles within 2° (including femur and tibia coronal and sagittal angles) (Figure 4). Another series of 306 TKA procedures done by one experienced surgeon using standard instrumentation showed only 52% alignment reproducibility.19 These results confirmed the current meta-analyses from Bauwens17 and Mason18 demonstrating the potential benefit of navigation in alignment. Only long-term follow-up will indicate whether improving alignment increases implant survivorship. However, despite the short-term review, we found a few radiolucent lines less than 1 mm, with no loosening, which corroborate leg alignment improvement.23 Most of these lines were not on the tibia but on the posterior femur, which was attributable to insufficient posterior cement (Figure 3).

The overall results of this extensive 1-year review of 214 TKAs are encouraging. These patients will be reviewed at 2, 5, and 10 years in our arthroplasty department, and we will report on long-term Columbus-navigated TKAs and verify whether navigation improved implant survival. We confirmed that using a nonimage-based navigation system routinely does not affect our practice, the number of complications, or our outcomes.

References:

  1. Callahan CM, Drake BG, Heck DA, et al. Patient outcomes following tricompartmental total knee replacement. A meta-analysis: Postoperative alignment of total knee replacement. Its effect on survival. JAMA. 1994;271(17):1349-1357.
  2. Kane RL, Saleh KJ, Wilt TJ, et al. Total Knee Replacement. Evid Rep Technol Assess (Summ). 2003;(86):1-8.
  3. Saleh KJ, Mulhall KJ, Hofmann AA, et al. Primary Total Knee Arthroplasty Outcomes. In: Barrack RL, Booth RE, Lonner JH, eds. Orthopaedic Knowledge Update: Hip and Knee Reconstruction 3. AAOS; 2006:93-110.
  4. Lizaur AL, Marco L, Cebrian R. Preoperative factors influencing the range of movement after total knee arthroplasty for severe osteoarthritis. J Bone Joint Surg Br. 1997; 79(4):626-629.
  5. McPherson EJ, Cushner FD, Schiff CF, et al. Natural history of uncorrected flexion contractures following total knee arthroplasty. J Arthroplasty. 1994; 9(5):499-502.
  6. Ritter MA, Lutgring JD, Davis KE, et al. The role of flexion contracture on outcomes in primary total knee arthroplasty. J Arthroplasty. 2007; 22(8):1092-1096.
  7. Walker PS. Total knee implant design. In: Barrack RL, Booth RE, Lonner JH, eds. Orthopaedic Knowledge Update: Hip and Knee Reconstruction 3. AOS Orthopaedic Knowledge Update. AAOS; 2006:31-40.
  8. Lombardi AV. Orthopaedic Secrets. In: Complications in Total Knee Arthroplasty. Philadelphia, Pa: Hanley and Belfus; 1999:306-308.
  9. Stöckl B, Nogler M, Rosiek R, et al. Navigation improves accuracy of rotational alignment in total knee arthroplasty. Clin Orthop Relat Res. 2004; (426):180-186.
  10. Chauhan SK, Scott RG, Breidahl W, et al. Computer assisted knee arthroplasty versus a conventional jig-based technique. J Bone Joint Surg Br. 2004; 86(3):372-377.
  11. Siston RA, Patel JJ, Goodman SB, et al. The variability of femoral rotational alignment in TKA. J Bone Joint Surg Am. 2005; 87(10):2276-2280.
  12. Picard F, Gregori A, Dean F, et al. Computer-assisted dynamic total knee arthroplasty using Whiteside’s line for alignment [erratum in: Orthopedics. 2007; 30(4):327]. Orthopedics. 2006; 29(10 suppl):S104-S107.
  13. Katz JN, Barret J, Mohamed NN, et al. Association between hospital and surgeon procedure volume and the outcomes of TKR. J Bone Joint Surg Am. 2004; 86-A(9): 1909-1916.
  14. Katz JN, Bierbaum BE, Losina E. Case mix and outcomes of TKR in orthopaedic specialty hospital. Med Care. 2008; 46(5):476-480.
  15. SooHoo NF, Lieberman JR, Ko CY, et al. Factors predicting complication rates following total knee replacement. J Bone Joint Surg Am. 2006; 88(3):480-485.
  16. Siston RA, Giori NJ, Goodman SB, et al. Surgical navigation for total knee arthroplasty: a perspective. J Biomech. 2007; 40(4):728-735.
  17. Bauwens K, Matthes G, Wich M, et al. Navigated total knee replacement. A meta-analysis. J Bone Joint Surg Am. 2007; 89(2):261-269.
  18. Mason JB, Fehring TK, Estok R, et al. Meta-analysis of alignment outcomes in computer-assisted total knee arthroplasty surgery. J Arthroplasty. 2007; 22(8):1097-1106.
  19. Lerat JL, Besse JL, Desmé D, et al. Accuracy of total knee replacement bone cuts using a conventional ancillary system: 300 Innex total knee arthroplasties [in French]. Rev Chir Orthop Reparatrice Appar Mot. 2006; 92(3):248-256.
  20. Sundaram RO, Parkinson RW. Closed suction drains do not increase the blood transfusion rates in patients undergoing total knee arthroplasty. Int Orthop. 2007; 31(5):613-616. Epub 2006 Sep 1.
  21. Kalairajah Y, Simpson D, Cossey AJ, et al. Blood loss after total knee replacement: effects of computer-assisted surgery. J Bone Joint Surg Br. 2005; 87(11):1480-1482.
  22. Jenny JY, Clemens U, Kohler S, et al. Consistency of implantation of a total knee arthroplasty with a non-image-based navigation system: a case-control study of 235 cases compared with 235 conventionally implanted prostheses. J Arthroplasty. 2005; 20(7):832-839.
  23. Jeffery RS, Morris RW, Denham RA. Coronal alignment after total knee replacement. J Bone Joint Surg Br. 1991; 73(5):709-714.
  24. Sikorski JM, Blythe MC. Learning the vagaries of computer-assisted total knee replacement. J Bone Joint Surg Br. 2005; 87(7):903-910.
  25. Callaghan JJ, Liu SS, Warth LC. Computer-assisted surgery: a wine before its time: in the affirmative. J Arthroplasty. 2006; 21(4 suppl 1):27-28.
  26. Unitt L, Sambatakakis A, Johnstone D, Briggs TW. Short-term outcome in total knee replacement after soft-tissue release and balancing. J Bone Joint Surg Br. 2008; 90(2):159-165.
  27. Peersman G, Laskin R, Davis J, Peterson M. The Insall Award paper - Infection in total knee replacement - A retrospective review of 6489 total knee replacements. Clin Orthop Relat Res. 2001; (392):15-23.

Authors

Drs Katipalli, Deakin, Greaves, Reynolds and Picard are from the Department of Orthopaedics, Golden Jubilee National Hospital, Clydebank, Glasgow United Kingdom.

Dr. Picard has a patent license from B. Braun Aesculap. Dr. Deakin’s department receives a research grant from B. Braun Aesculap. Drs Katipalli, Greaves, and Reynolds have no relevant financial relationships to disclose.

Correspondence should be addressed to: Frederic Picard, MD, Department of Orthopaedics, Golden Jubilee National Hospital, Clydebank, Glasgow, G81 4HX, United Kingdom.