The present IRB approved study evaluates the early results of 100 TKAs using CT-based Patient-Specific Instrumentation (PSI) (MyKnee®, Medacta International, SA, Castel San Pietro, Switzerland). For this technique, a CT scan of the lower extremity is obtained, and from these images, the knee is reconstructed 3-dimensionally. Surgical and implant-size planning are performed according to surgeon preference, with the goal to create a neutral mechanical axis. Once planned and approved, the blocks are made.

Outcomes measured for the present study include surgical factors such as Tourniquet Time (TT) as a measure of surgical efficiency, the actual intraoperative bony resection thicknesses to be compared to the planned resections from the CT scan, and complication data. Furthermore, pre- and post-operative long standing alignment and Knee Society Scores (KSS) were obtained.

During surgery, the PSI cutting block is registered on the femur first and secured with smooth pins. No osteophytes are removed as the blocks use the positive topography of the osteophytes for registration. The distal femoral resection is performed directly through the block. An appropriate sized 4-in-1 block is placed and the remaining resections are performed. The tibial resection block is registered and resection performed. Final bone preparation, patella resurfacing, and trialing is performed as is standard to all surgical techniques.

There were 50 Left and 50 Right TKA's performed in 61 females and 39 males. All patients had diagnosis of osteoarthritis. The average BMI was 31.1 and average age was 64.5 (range 41–90). 79 patients had pre-operative varus deformities with Hip Knee Angle (HKA) average of 174.7° (range 167°–179.5°). 19 patients had pre-operative valgus deformities averaging 184.4° (range 180.5°–190°). Three patients were neutral.

Average TT was 31.2 minutes (range 21–51 minutes). With regard to the bony resections, the actual vs. planned resections for the distal medial femoral resection was 8.7 mm vs. 8.9 mm respectively. Further actual vs. planned femoral resections include distal lateral 7.2 vs. 6.7 mm; posterior medial 8.3 vs. 8.9 mm; and posterior lateral 6.2 vs. 6.8 mm. The actual vs. planned tibial resections recorded include medial 6.4 vs. 6.3 mm and lateral 8.3 vs. 8.2. The planned vs. actual bony cuts are strongly correlated, and highly predictive for all 6 measured cuts (p=<.001). No intraoperative complications occurred.

Average KSS improved from 45.9 to 81.4, and KSS Function Score improved from 57.7 to 73.5 at 6 weeks postoperative visit. There were no thromboembolic complications. Two patients had a post-operative infection requiring surgical intervention.

Post-operative alignment was 179.36° (range 175°–186°) for all patients. Alignment was neutral, within 3° in 95.9% of patients. There were only 4 outliers with maximal post-operative angulation of 6°.

In conclusion, these early results demonstrate efficacy of CT-based PSI for TKA. The surgery can be performed efficiently, accurately, and safely. Furthermore, excellent short term clinical and radiographic results can be achieved.

Hardware in or about the knee joint presents a number of challenges to the surgeon in performance of Total Knee Arthroplasty (TKA). Conventional instrumentation usually requires a modification of technique or removal of the metallic implants. Computer-Assisted TKA (CAOS) is another option, but adds complexity and time to the procedure. MRI-based Patient-Specific Instrumentation (PSI) cannot be used as metal causes unwanted artifact and renders the images for planning, useless. However, CT scans are not affected by metal and thus CT-based PSI can be used in TKA patients with pre-existing hardware.

The present IRB approved study evaluates 12 consecutive knees (10 patients) with pre-existing hardware using CT-based PSI (MyKnee®, Medacta International, SA, Castel San Pietro, Switzerland).

In this technique, CT scan of the lower extremity is obtained, and from these images, the knee is reconstructed 3-dimensionally. Surgical and implant-size planning are performed according to surgeon preference, with the goal to create a neutral mechanical axis. Once planned and approved, the blocks are made.

During surgery, the PSI cutting block is registered on the femur first and secured with smooth pins. The distal femoral resection is performed directly through the block. An appropriate sized 4-in-1 block is placed and the remaining femoral resections are performed. The tibial resection block is registered and resection performed. Final bone preparation, patella resurfacing, and trialing is performed as is standard to all surgical techniques.

Of the 12 TKAs, there were 5 left and 7 right knees performed in 6 females and 6 males. The average BMI was 33.19 and average age was 53 (range 44–63). All diagnoses were either osteoarthritis or post-traumatic osteoarthritis. Follow-up averaged 59 weeks (range 18.6–113.7).

Nine patients had pre-operative varus deformities with HKA deformities average of 171.9° (range 154°–178.5°). One patient had pre-operative valgus deformity of 184.5°. Two patients were neutral (180°). Post-operative alignment for all patients (n=11) was 179° (range 177°–180°). All patients were within 3° neutral, post operatively. Four patients measured 180°, 4 measured at 179°, 2 measured at 178°, and only one at 177°.

Hardware consisted of 5 patients with femur or tibia staples, 3 with plate(s) and screws, 3 patients with ACL interference screws, and one titanium rod. No hardware was removed unless necessary for implantation. Only 3 patients required some hardware removal.

The pre-operative Range of Motion (ROM) averaged 2.9° to 98.3° (Extension range 0–15° and flexion range 30–115°). Post-operative ROM was 2.9° to 101.3°. (Extension range 0–5° and flexion range 65–125°). Knee Society Score (KSS) improved from 42.3 to 82.3, and KSS Function Score improved from 52.1 to 77.5. No intraoperative complications were recorded. Average tourniquet time was 42.1 minutes (range 28–102).

Regardless of the deformity, the patient's post-operative mechanical axes HKA averaged 179° (range 177–180). Clinical scores were typical for TKA patients with improvement in both KSS and ROM.

In conclusion, early results using PSI in patients with pre-existing hardware in or about the joint, is safe, efficient, and accurate in performance of TKA.

In May 2010, MyKnee® patient-specific instrumentation was approved for use in this procedure in the USA. This technique uses a pre-operative CT scan of the lower extremity to plan the surgery. Images of the hip, knee, and ankle are reconstructed digitally to assess pre-operative deformity as well as size of the knee. Surgery is then planned with the goals of restoring a neutral mechanical axis of limb and providing correct sizing and placement of implants after the surgery. From this plan, patient-specific jigs are created to perform the surgery achieving the planned result without sacrificing speed of surgery or increasing complexity of the procedure. The present study seeks to evaluate both intraoperative and radiographic results of this procedure. IRB approval for retrospective research was obtained prior to evaluation of the data.

Thirty consecutive patients (14 males, 16 females) underwent TKA using the MyKnee technique by the senior author. Pre-operative long-standing radiographs were taken and compared to 6-week post-operative radiographs. Intraoperative data includes the femoral and tibial resection thickness: distal medial femoral, distal lateral femoral, posterior medial femoral, posterior lateral femoral, medial tibia, and lateral tibia. These were compared to the planned vs. actual resections. Tourniquet time was recorded as a measure of speed of surgery. These were compared to 30 consecutive patients using standard TKA technique by the same author. Intraoperative complications were also recorded.

For patients with varus pre-operative deformities (n = 21), the mechanical alignment was 7.8° (range 1.2° to 15.2°). Seven patients had pre-operative valgus deformities averaging 6.93° (range 1.3° to 14.5°). Two patients were neutral. Post-operative alignment for all patients (n = 30) was varus 1.92° (range 0° to 5.8°). Seventy-eight percent of patients were within 3° and 97% of patients were within 3.6°. In comparison, post-operative alignment for standard TKA patients measured varus 1.85°, which was not statistically significant. Seventy-nine percent of patients were within 3°; however the outliers were more dramatic ranging 3.5° to 9.2°.

Thirty femoral and 21 tibial resections were available for review using the MyKnee technique. The actual vs. planned resections for the distal medial femoral resection was 9.5 vs. 9.1mm respectively. Further actual vs. planned femoral resections include distal lateral femoral 8.4 vs. 6.3mm; posterior medial femoral 9.3 vs. 9.5mm; and posterior lateral femoral 8.6 vs. 7.0mm. The actual vs. planned tibial resections recorded include medial 6.07 vs. 6.29mm and lateral 9.36 vs. 8.19mm. Tourniquet time averaged 32.97 minutes (range 25 to 54) in the standard TKA group vs. 37.03 minutes (range 1 to 71) in the MyKnee group. This difference was not significant. However, the final 15 MyKnee patients had an average time of 33.46 minutes. No intraoperative complications occurred.

Many techniques exist for performance of TKA. Patient-specific cutting blocks allow the surgeon to pre-operatively determine resection depths, rotations, alignment, and sizing prior to the operative procedure itself. The present study shows that intraoperative resections and post-operative alignments can be accurately achieved with pre-operative CT planning and using patient-specific instrumentation.

For the typical varus knee deformity, cartilage will exist on the lateral side of the knee. This can cause measurement error when measuring the lateral compartments as the CT scan is based on bone only. This can be seen in 2.1mm and 1.6mm differences in the distal lateral femoral and posterior lateral femoral resections respectively. Thus, this difference can be explained by the false measurement of intact cartilage. More accurate results could be obtained if the cartilage was removed and bone measured. Valgus knees, being diseased in the lateral compartment, did not show such variance as expected in planned vs. actual resections. Intraoperative speed of surgery is important to all participants in TKA: surgeon, hospital, and patient. Obviously accuracy should not be sacrificed for speed so it is important for any new technology introduced to the market to accelerate surgery not compromise results. In the current study, the average times of MyKnee vs. standard TKA surgery were comparative and not significantly different using a two-sample T-test. The standard TKA average tourniquet time may appear faster than other reported literature; however the surgeon is on the end of learning curve with the system. The MyKnee average tourniquet time represents the initial procedures in the learning curve and can be considered slower than what they will eventually be as the author gains more experience with the technique. Efficiency was demonstrated with the decrease in tourniquet time for the last 15 patients.

Furthermore, the goals of surgery were maintained radiographically. Regardless of the deformity, the patient's post-operative mechanical axes averaged 1.85° for standard technique and 1.92° for the MyKnee group, not statistically significantly different. These results were obtained via long-standing x-rays, which are well known to be prone to error in alignment secondary to potential flexion and rotation of the extremity. The standardised protocol for acquisition of the X-ray, attempts to prevent these errors and X-rays are routinely re-done if the technician feels error has occurred. The technique also appears safe as no intra-operative complications occurred and were recognised within the first six weeks post-operative.

In conclusion, using patient-specific instrumentation (MyKnee) is safe, quick, and accurate in performance of TKA.