Advancements in Minimally Invasive Total Knee Arthroplasty

ABSTRACT

Alfred J. Tria Jr, MD

Total knee arthroplasty (TKA) has been in development since the early 1970s. Insall and others established the principles of ligament balance and overall alignment for implant success. Repicci introduced the concept of minimally invasive surgery in the early l990s using the unicondylar prosthesis. As the outcomes of minimally invasive surgeries continued to improve when using a unicondylar prosthesis, it was logical to attempt a minimally invasive TKA. The author and his team have performed 120 minimally invasive TKAs over the past 2 years. Early results show that a minimally invasive approach produces better early motion, less blood loss, less pain, and a shorter hospital stay than the standard TKA with no compromise in accuracy.

Designers of total knee arthroplasty (TKA) were initially concerned with ligament balancing and knee alignment.¹ Exposure was of key importance to evaluate the TKA and confirm the position of the components.

During the 1970s, the surgical technique was improved and refined. Insall et al² introduced the concept of ligament release and Krackow et al³ described ligament tightening techniques. Instruments were created to support the surgical approaches and referenced both the intramedullary canal and extramedullary landmarks.^4,5 Knee implants underwent considerable modification during the late 1970s and 1980s. The total condylar knee gave way to the total condylar prosthesis type II⁶ and then to the posterior stabilized knee.⁷ The porous-coated anatomic knee introduced cementless fixation and supported the posterior cruciate retaining designs,⁸ and the mobile- bearing design was introduced in the late 1970s and has been successful since its introduction.⁹

Unicondylar knee arthroplasty began in the early 1970s at the same time that the total condylar knee was being designed.^10,11 The enthusiasm for this implant system was tempered by the early failures due to design issues and surgical technique. Marmor’s¹² fixed-bearing unicondylar knee arthroplasty and the Oxford mobile-bearing unicondylar knee arthroplasty¹³ were two major implants that were developed simultaneously. In the late 1980s, Miller and Galante introduced their fixed-bearing design along with improved surgical techniques created specifically for the implant.¹⁴ They did not seek to overcorrect the knee. They avoided ligament releases and intentionally left the knee slightly more lax than the comparable TKA. Their follow-up is now approaching 15 years with >90% survival.¹⁴

It was not until the 1990s that Repicci et al^15,16 introduced minimally invasive surgery based on unicondylar knee arthroplasty. They decreased the surgical incision and encouraged surgeons to recognize the differences between unicondylar knee arthroplasty and TKA. As minimally invasive surgical techniques for unicondylar knee arthroplasty improved, surgeons began to question if the principles of standard TKA could be applied to minimally invasive TKA.

Materials and Methods

Thomas M. Coon, MD (Shasta Orthopaedics and Sports Medicine, Red Bluff, Calif) and E. Marlowe Goble, MD (Medicine Lodge, Logan, Utah) are two principal investigators. Initially, Coon designed extramedullary instruments for minimally invasive unicondylar knee arthroplasty (Figure 1). The instruments attached to the extramedullary tibial cutting guide and required distal femoral condyle resection from the side of the knee. Although the approach was novel, surgeons found it difficult to adjust. During the course of his work with the instruments, it became evident that similar techniques could be applied to TKA for initial minimally invasive TKA. Coon’s extramedullary instruments were modified. However, modification of the existing instrumentation would not permit minimally invasive TKA, and the author’s team designed completely new instruments.

Figure 1: Extramedullary rods are used to reference flexion-extension and varus-valgus of the distal femoral cut in unicompartmental knee arthroplasty.	Figure 2: The medial incision begins from the superior pole of the patella and extends just slightly below the tibiofemoral joint line. The solid line A is the joint line. Line B is the outline of the medial femoral condyle (A). The lateral incision is slightly more vertical than the medial but also begins at the superior pole of the patella and extends just below the joint line (B).

To limit the variables at the start of this study, the authors implanted the NexGen LPS Flex knee (Zimmer, Warsaw, Ind) in all patients. This knee is a posterior stabilized design with an increased thickness for the posterior femoral runner and an increased lip on the posterior aspect of the tibial polyethylene insert. The design favors increased motion and resects 2 mm more bone from the posterior aspect of the femoral condyles than the Legacy Posterior Stabilized prostheses (Zimmer). The slightly greater bone resection increases the flexion gap space and facilitates the minimally invasive surgical procedure.

The surgeons use an 8-cm incision that begins at the superior pole of the patella on the medial side and ends 2 cm distal to the tibial joint line (Figure 2). The arthrotomy is made in line with the skin incision and does not violate the quadriceps tendon or the vastus medialis muscle; it is purely a capsular incision. The skin incision may be slightly longer to accommodate a tall or heavy patient. The length of the skin incision does not define minimally invasive surgery. The management of the quadriceps tendon and surrounding muscles is the defining feature. The patellar fat pad is excised to increase the view of the joint, and the patella is resurfaced.

Resurfacing is preferably performed with instrumentation to increase the accuracy, but it can also be performed free hand while referencing the appropriate bony landmarks on the patellar surface. The thickness of the patellar surface should be decreased by approximately 2 mm from the initial thickness. Resection of the patellar surface increases the available space in the knee joint.

The distal femoral cutting guide requires intramedullary referencing and allows for a distal femoral valgus of 4°, 6°, or 8° (Figure 3). Cutting both condyles from the medial side is somewhat awkward for an inexperienced surgeon; however, with practice the cuts can be made safely and accurately. The tibial cutting guide is extramedullary and the cutting block is placed around the medial side of the tibia for greater accuracy and ease (Figure 4). Removal of the proximal tibial bone is a difficult step of the procedure.

Figure 3: The intramedullary femoral cutting guide references the medial femoral condyle (A). A cutting block attaches to the guide and directs the saw blade across the distal femoral condyles from medial to lateral (B).		Figure 4: The tibial cutting guide extends around the medial side of the tibial plateau and a depth gauge is used to establish proper thickness of the cut.

It is important to control the saw blade when penetrating the posterior and lateral cortex of the tibia to protect surrounding soft-tissue structures. The author has had one experience where the middle geniculate was cut in the posterior aspect of the knee and the incision had to be abandoned to control the bleeding vessel. Although this knee had no resultant ill effects, it was a warning of the dangers involved. Goble (A.J.T., personal communication) suggested that the procedure begin with arthroscopic surgery to release all of the soft tissue about the tibia and to resect the cruciate ligaments on the femoral and tibial sides. Presently, the author’s team is implanting only posterior stabilized total knees but there are plans to incorporate the cruciate retaining designs.

After the horizontal tibial cut is completed, a sagittal cut is made in the intercondylar notch and the medial one-half of the tibial bone is removed to allow better visualization of the lateral aspect. After removing the remainder of the tibial bone, the extension space is sized and the collateral ligaments balanced. If ligament releases are required, both the medial and lateral sides of the knee are readily visible because 20 mm of bone is removed from the proximal tibia and distal femur.

The finishing cuts for the distal femur were initially difficult because the author attempted to modify the original cutting blocks. The instruments were redesigned to coordinate with the smaller incision and the medial surgical approach. The new guide uses posterior condylar referencing and incorporates 3° of external rotation for the cutting block (Figure 5). The tower of the guide references the anteroposterior (AP) axis of Whiteside. If one of the condyles is atrophic, the guide must be adjusted by moving the posterior pad of the guide posterior to the affected condyle. A guide rail slides beneath the cutting guide and incorporates 3° of external rotation. When the appropriate cutting block is locked onto the rail, the AP position is predetermined. After correcting the medial lateral position in full extension, all of the cuts can be completed at once. The flexion and extension gaps can now be compared and adjusted as necessary. The box cut is made on the femoral side after balancing and alignment are confined for the final time.

Figure 5: The femoral cutting guide has two pads for the femoral condyles and a boom to reference the anterior cortex (A). The femoral cutting guide is shown in position in the knee (B).		Figure 6: The tibial finishing guide is positioned on the surface of the tibial plateau.

The tibial sizing and positioning are still slightly difficult at this time because the lateral aspect of the tibia is not readily visible (Figure 6). Pins are added to the posterior part of the tibial finishing guide and the anterior handle is changed to avoid impingement on the patellar tendon. The guide is lined up with the tibial tubercle, the box cut of the femur, and the malleoli of the ankle in the same fashion as the standard TKA tibial tray.

The trial components are tested to ensure that range of motion is full, the patella tracks in the midline, the gaps are balanced, and that the overall valgus alignment is restored. The surfaces are lavaged and cleared for cementing and all of the components are cemented. The tibial tray is inserted first because it is difficult to insert the single-piece component with the intramedullary stem. The cruciate-retaining design with shortened base pegs should be easier to insert. The patella is cemented and subluxed laterally against the lateral femoral condyle wall. There is not enough room in the knee to place a compression clamp and leave it in position during the remainder of the cementing. The femur is cemented last and, sometimes, the skin incision must be slightly lengthened if the component is one of the two largest sizes.

Two surgical drains are placed into the knee joint and the tourniquet is released. The closure is performed in the standard fashion so that the medial capsule is closed tightly. A knee immobilizer or bulky dressings are not applied. During the surgery, a cell saver is used and the salvaged blood is returned to the patient in the postanesthesia care unit over a 1- to 2-hour period.

Patients begin ambulation within 2 hours postoperatively. The anesthesia team should be informed about this early physical therapy so the anesthetic (either endotracheal general or spinal-epidural) can be planned appropriately. The majority of the author’s patients receive general anesthesia and this has not caused any complications with postoperative management.

The drains are removed the morning after surgery and patients remain in the hospital for 48 hours. They are subsequently transferred to an acute care rehabilitation center for 5 days of inpatient therapy. The typical TKA patient begins physical therapy on the first postoperative day, remains hospitalized 4 days postoperatively, and is transferred to acute rehabilitation for 7 days. Intravenous (IV) antibiotics are given for 24 hours postoperatively.

The risk for venous thrombosis persists despite advancements in surgical technique. Without prophylaxis, the rate of venous thromboembolism after TKA is high. Calf deep vein thrombosis (DVT) can occur in 57% to 71% of patients, and proximal DVT has been reported in 10% to 23% of patients.¹⁷

Pharmacologic prophylaxis with warfarin, unfractionated heparin, and low-molecular-weight heparin have been the mainstay of venous thromboembolism prophylaxis.

Though effective, these agents do not completely eliminate the risk for venous thromboembolism following TKA (Table 1). Fondaparinux, a synthetic factor Xa inhibitor, has demonstrated a significant reduction in venous thromboembolism risk compared with the low-molecular-weight heparin, enoxaparin, with no differences in the incidence of clinically relevant bleeding.¹⁸ Timing of fondaparinux after surgery is critical in allowing for optimum balance between efficacy and safety, with the first dose given no sooner then 6-8 hours postoperatively.

For the last 22 years, the author has used low dose warfarin in >5000 TKA patients with a good success rate. However, warfarin is a narrow therapeutic index drug that requires frequent monitoring of its anticoagulant effect to optimize its efficacy and minimize bleeding complications. Warfarin has limitations due to drug–drug and drug–food interactions, requiring constant patient screening. Although lengths of stay decrease with minimally invasive surgery, warfarin still requires systematic, frequent follow-up of patients for monitoring and dose adjustments. Without a good support system, such as an established anticoagulation monitoring service, the use of warfarin can lead to complications.¹⁹

Alternative options include newer agents that do not require dose adjustments or monitoring of the anticoagulant effect. Fondaparinux meets these criteria and was recently approved as thromboprophylaxis in patients undergoing minimally invasive surgery at the author’s institution. Table 2 describes the venous thromboembolism prophylaxis protocol applied by the author. Early results in 22 patients who received unicondylar knee arthroplasty demonstrated negative Doppler at 10 days postoperatively in all patients with no bleeding episodes or wound complications.

Results

The results presented in this section are early findings and should not be compared to published results with 2-15 years’ follow-up. At this time, only 15 patients are beyond 6 months postoperatively and only three patients are beyond 1 year postoperatively (the initial operation was performed in February 2002).

The author has attempted 60 minimally invasive TKA procedures. Two knees were converted to a standard TKA. One knee was changed because of soft bone in an obese patient with rheumatoid arthritis and the second was changed because of posterior capsular bleeding from the middle geniculate artery. Thus, 58 knees are presently available for study (average age: 67 years; range: 51-86 years). Eight patients underwent bilateral procedures performed sequentially under the same anesthetic. There are 29 women and 21 men. The average tourniquet time was 110 minutes and the shortest time was 75 minutes. The blood loss by cell saver tabulation is 200 cc (compared to the standard TKA loss of 350-400 cc). The pain score at physical therapy on the first visit averaged 5.2 compared to 7.5 for standard TKA. The hospital length of stay is 2.5 days versus 4 days for standard TKA.

Complications include one nonfatal pulmonary embolism, one intraoperative nonfatal myocardial infarction that subsequently led to a right-sided hemiparesis that is slowly resolving, and two transient postoperative arrhythmias. There have been no infections or skin compromises.

Discussion

Data were collected from 20 previous TKAs performed using a NexGen LPS Flex fixed-bearing knee design during 18 months. There were no bilateral procedures in this series and patients were chosen for the implant because of unusually high preoperative range of motion. The patients underwent surgery just before the minimally invasive surgery TKA project began.

The average preoperative range of motion of standard TKA procedures was 130° and the range of the minimally invasive TKA group was 119°. At the first postoperative visit (2-4 weeks postoperatively), the standard knees had 91° of flexion and the minimally invasive surgical TKAs had 112°. At 6 weeks postoperatively, standard knees had 115° of flexion and minimally invasive TKAs had 126° flexion. This difference was not statistically significant but showed a trend toward greater motion with the minimally invasive TKA. Postoperative radiographs were measured and compared. The overall valgus of the postoperative knees in both groups was 4°. The tibial cut in the minimally invasive TKA was in 2.5° of varus versus 1° of varus for the standard knee. This difference was not significant but also showed a trend. The tibial instrument was changed to address this difference.

The preliminary data suggest that patients with minimally invasive TKAs obtain motion faster than patients receiving standard TKA. It also suggests that the tibial cut in the minimally invasive TKA is in slight varus, probably secondary to the long tibial blade coming across the tibia from medial to lateral. However, the overall accuracy of minimally invasive TKA appears to be similar to standard TKA.

Conclusion

The criticism of minimally invasive TKA revolves around the lack of long-term results, controls, peer-reviewed publications, and the use of the technology for personal aggrandizement. The early results that are reviewed in this paper support the fact that the procedure has positive results immediately postoperatively. The technical quality of the arthroplasty is equal to that of standard TKA.

Peer-reviewed articles will most likely follow when 2-year follow-up is available and appropriate comparisons need to be drawn. However, in a modern society with rapid, worldwide transmittal of information, it would not be appropriate to hide the technology or prohibit its growth.

The author has studied this surgical procedure in a methodical manner. The instruments have been refined and the procedure can be performed with accuracy equal to standard TKA. The next step is to design a new TKA that can be implanted in a modular fashion to facilitate the surgical procedure. The TKA will be a modification of the NexGen Legacy Posterior Stablized (Zimmer) Flex fixed-bearing implant. We must show that the knee is user friendly and can produce long-term results comparable to the present LPS line of knees. Navigational computerized equipment is included with the new line of instruments and is hoped to improve clinical results.

References

1. Insall J, Ranawat CS, Scott WN, Walker P. Total condylar knee replacement: preliminary report. Clin Orthop. 1976; 120:149-154.
2. Insall J, Tria AJ, Scott WN. The total condylar knee prosthesis: the first 5 years. Clin Orthop. 1979; 145:68-77.
3. Krackow KA, Thomas SC, Jones LC. A new stitch for ligament-tendon fixation. Brief note. J Bone Joint Surg Am. 1986; 68:764-766.
4. Bertin KC. Instrumentation. In: Fu FH, Harner CD, Vince KG, eds. Knee Surgery. Baltimore, MD: Williams and Wilkins; 1994:1303-1345.
5. Engh GA, Petersen TL. Comparative experience with intramedullary and extramedullary alignment in total knee arthroplasty. J Arthroplasty. 1990; 5:1-8.
6. Insall JN, Tria AJ. The total condylar prosthesis type II. Paper presented at: American Academy of Orthopaedic Surgeons. Annual meeting. San Francisco, Calif; 1979.
7. Stern SH, Insall JN. Posterior stabilized prosthesis. Results after follow-up of nine to twelve years. J Bone Joint Surg Am. 1992; 74:980-986.
8. Hungerford DS, Krackow KA. Total joint arthroplasty of the knee. Clin Orthop. 1985; 192:23-33.
9. Buechel FF Sr. Long-term followup after mobile-bearing total knee replacement. Clin Orthop. 2002; 404:40-50.
10. Marmor L. The modular knee. Clin Orthop. 1973; 94:242-248.
11. Jones WT, Bryan RS, Peterson LFA, Ilstrup DM. Unicompartmental knee arthroplasty using polycentric and geometric hemicomponents. J Bone Joint Surg Am. 1981; 63:946-954.
12. Marmor L. Unicompartmental arthroplasty of the knee with a minimum 10-year follow-up period. Clin Orthop. 1988; 228:171-177.
13. Murray DW, Goodfellow JW, O’Connor JJ. The Oxford medial unicompartmental arthroplasty: a ten-year survival study. J Bone Joint Surg Br. 1998; 80:983-989.
14. Berger RA, Nedeff DD, Barden RM, et al. Unicompartmental knee arthroplasty. Clinical experience at 6- to l0-year follow up. Clin Orthop. 1999; 367:50-60.
15. Repicci JA, Eberle RW. Minimally invasive surgical technique for unicondylar knee arthroplasty. J South Orthop Assoc. 1999; 8:20-27.
16. Romanowski MR, Repicci JA. Minimally invasive unicondylar arthroplasty: eight-year follow-up. J Knee Surg. 2002; 15:17-22.
17. Geerts WH, Heit JA, Clagett GP, et al. Prevention of venous thromboembolism. Chest. 2001; 119:132S-175S.
18. Turpie AG, Bauer KA, Eriksson BI, Lassen MR. Fondaparinux vs enoxaparin for the prevention of venous thromboembolism in major orthopedic surgery: a meta-analysis of 4 randomized double-blind studies. Arch Intern Med. 2002; 162:1833-1840.
19. Ansell J, Hirsh J, Dalen J, et al. Managing oral anticoagulant therapy. Chest. 2001; 119:22S-38S.

Authors

From Robert Wood Johnson Medical School, New Brunswick, NJ.