人工膝関節全置換術における患者特異的切断ガイド

Accurate alignment of components is considered important for improved outcomes when restoring lower limb alignment in total knee arthroplasty (TKA). Patient-specific guides were developed with the intention of improving surgical efficiency, accuracy of component positioning and overall limb alignment. Currently, the benefits of patient-specific guides in achieving these goals and whether this has an impact on clinical and functional outcomes remains unclear.

To assess the benefits and harms of patient-specific cutting guides versus conventional cutting guides or computer-assisted surgical navigation in people undergoing primary TKA.

We searched the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE (Ovid), Embase (Ovid) and trial registers up to 21st January 2025, unrestricted by language.

We included randomised controlled trials (RCTs) comparing patient-specific cutting guides (PSGs) to conventional instrumentation (CON) or computer-assisted surgical navigation (NAV) in TKA. Major outcomes were survival of implant (risk of revision), function, radiographic lower limb alignment, pain, global assessment, total adverse events and re-operation rate.

We used standard methods recommended by Cochrane.

Forty-four studies with 3664 participants were identified. Out of these, 40 studies with 3134 participants compared PSG to CON, two studies with 140 participants compared PSG to NAV and two studies with 390 participants compared PSG to NAV and CON. Regarding the imaging modality, 27 trials used magnetic resonance imaging (MRI)-based PSGs, 18 trials used computed tomographic (CT- based PSGs and 2 trials used both CT-based and MR-based PSGs). The mean age of participants ranged from 63 years to 74 years. Fifty-three per cent of participants were male, with more than 90% of participants having knee osteoarthritis. Most of the included trials were at risk of bias; 25/44 (56.8%) studies were at risk of selection bias, and 39/44 (88.6%) were at risk of performance and detection biases. 1. PSGs compared to conventional instrumentation Compared to conventional instrumentation, PSGs may result in little to no difference in survival of the implant. At 26 months, 7/347 participants (20 per 1000) in the conventional instrumentation group reported survival of implant (risk of revision) compared to 4/342 participants (16 per 1000) in the PSG group (RR 0.79 (95% CI 0.25 to 2.52); I² = 0%; 689 participants; 9 studies; low-certainty evidence downgraded for bias and imprecision). Compared to conventional instrumentation, PSGs may result in little to no improvement in function. Mean function (KSS, OKS, WOMAC, KOOS) was 15.1 points with conventional instrumentation and 13.44 points with PSG (SMD -0.11, 95% CI -0.25 to 0.03 back-translated to MD 1.66 points lower (95% CI 3.77 points lower, 0.45 points higher; I² = 56%; 23 studies, 1913 participants; low-certainty evidence downgraded for bias and indirectness)) at short-term (up to two years). PSGs may result in little to no difference in precision in lower limb alignment as measured by the proportion of outliers for femorotibial coronal angle (FTCA). At up to two years, 266/1208 participants (220 per 1000) in the conventional instrumentation group reported radiographic outlier events compared to 234/1204 participants (189 per 1000) in the PSG group (RR 0.86, 95% CI 0.68 to 1.10, I² = 49%; 29 studies, 2412 participants; low-certainty evidence downgraded for bias and imprecision). Low-certainty evidence, downgraded for bias and indirectness, showed that PSGs may result in little to no difference in pain. Mean pain (0 to 100 scale, 0 no pain) was 17.4 points with conventional instrumentation and 15.8 points with PSG (SMD -0.09, 95% CI -0.23 to 0.06, back-translated to MD 1.52 points lower (95% CI 3.88 lower, 1.01 higher; I² = 0%, 10 studies, 715 participants)). None of the included studies provided data on global assessment. Low-certainty evidence, downgraded for bias and imprecision, showed that PSGs may result in little to no difference in adverse events. Total adverse events (infection, thrombosis) were reported in 60/590 participants (102 per 1000) in the conventional instrumentation group compared to 61/579 participants (99 per 1000) in the PSG group (RR 0.97, 95% CI 0.68 to 1.39, I² = 5%, 14 studies, 1169 participants). Low-certainty evidence, downgraded for bias and imprecision, showed that PSGs may result in little to no difference in re-operation rates. Re-operations were reported in 21/424 participants (50 per 1000) in the conventional instrumentation group compared to 16/419 participants (44 per 1000) in the PSG group (RR 0.87, 95% CI 0.45 to 1.68, I² = 0%, 10 studies, 843 participants). 2. PSGs compared to computer-assisted navigation Low-certainty evidence (downgraded for bias and imprecision) showed that, compared to computer-assisted navigation, PSGs may result in little to no difference in function. Mean function (0 to 100, 100 best function) was 52.5 points with computer-assisted navigation compared to 57.5 points with PSGs at the short term (less than two years): MD 5.00 points higher (95% CI 1.31 lower to 11.31 higher, I² = 0%, 2 studies, 120 participants). Low-certainty evidence (downgraded for bias and imprecision) showed that use of PSGs may result in worse precision in lower limb alignment compared to computer-assisted navigation. At the short term (less than two years), 13/135 participants (96 per 1000) in the computer-assisted navigation group reported radiographic outlier events compared to 28/135 participants (196 per 1000) in the PSG group (RR 2.04, 95% CI 0.87 to 4.82, I² = 44%, 3 studies, 270 participants). None of the included studies provided data on survival of the implant, global assessment, pain, re-operations or adverse events.

Low-certainty evidence showed no meaningful differences with regard to survival of implant, re-operation rate, adverse events, function, pain, and global assessment when comparing patient-specific cutting guides to conventional instrumentation or to computer-assisted navigation in total knee arthroplasty.