Temporal trends in forgotten joint score following high tibial osteotomy and unicompartmental knee arthroplasty: A systematic review and limited meta-analysis

INTRODUCTION

Medial compartment osteoarthritis accounts for nearly 50% of knee osteoarthritis cases and can be effectively managed with either high tibial osteotomy (HTO) or unicompartmental knee arthroplasty (UKA).^[1] Both procedures are well established in providing pain relief, improving symptoms, and enhancing post-operative functional outcomes. Conventionally, patient selection has been guided by age and activity level; however, contemporary evidence suggests that there is no strict boundary differentiating indications for HTO and UKA in patients with isolated medial compartment disease.

Several systematic reviews and meta-analyses have reported that UKA may be superior to HTO in terms of lower revision rates, reduced post-operative pain, fewer complications, and better functional outcome scores.^[2-4] Conversely, HTO has been shown to offer a relatively greater post-operative range of motion and is therefore often recommended for younger, more physically active patients. UKA, on the other hand, allows faster recovery and is commonly preferred in relatively older patients with lower functional demands.^[2-4]

From a conceptual standpoint, the two procedures differ fundamentally in their biomechanical principles. HTO addresses medial compartment overload by altering the lower limb alignment, thereby shifting the weight-bearing axis toward the lateral compartment, most commonly through a medial open-wedge osteotomy and less frequently through a lateral closing-wedge technique.^[5] In contrast, UKA involves resurfacing the medial tibiofemoral compartment without intentional alteration of overall limb alignment.^[6]

Given the alignment alteration inherent to HTO, patients may experience altered knee biomechanics and proprioceptive feedback in the post-operative period. This may result in increased joint awareness until central neuromuscular adaptation (adjustment of the brain) to the new alignment occurs.^[7] In addition, rehabilitation and recovery following HTO are typically longer compared to UKA. Consequently, it is plausible that patients undergoing HTO experience greater joint awareness than those treated with UKA, particularly in the early post-operative phase.

The literature, however, remains inconclusive regarding this aspect. While some studies have reported lower forgotten joint score (FJS-12) values following HTO compared to UKA, others have found no significant difference between the two procedures.^[8-12] Therefore, this systematic review and meta-analysis aims to compare FJS-12 scores at different post-operative time points between patients undergoing HTO and those treated with UKA, with the objective of clarifying differences in joint awareness between these two surgical strategies for medial compartment osteoarthritis. We hypothesize that joint forgetfulness will take longer time in patients undergoing HTO than in those treated with UKA.

MATERIALS AND METHODS

This systematic review and meta-analysis were conducted in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses guidelines and registered in the PROSPERO database before data collection (PROSPERO acknowledgment ID: 1303022).

RESULTS

Study identification, study quality, risk of bias, and patient demographics of the included studies

After screening, 26 articles were identified, of which five studies ultimately met the inclusion criteria and directly compared FJS-12 outcomes between UKA and HTO [Figure 1]. All included studies were retrospective comparative in design and reported clearly defined inclusion and exclusion criteria.^[8-12] Risk-of-bias assessment using the NOS demonstrated a low risk of bias in two studies, moderate risk of bias in one study, and high risk of bias in the remaining two studies [Table 1]. A total of 535 patients were included, comprising 286 UKA patients and 249 HTO patients. The overall mean age, calculated from studies reporting numerical age data, was approximately 60 years. The mean age in the UKA group was 65.3 years, whereas the mean age in the HTO group was 55.6 years, reflecting the younger patient profile typically selected for osteotomy [Table 1].

Close

Figure 1:

Flowchart showing study recruitment in this systematic review (as per the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines). FJS: Forgotten joint score-12.

Export to PPT

Table 1: Patient demographics in the reviewed studies.

Study	Study design	Participants	Sample size (UKA/HTO)	Age (years)	Gender (M:F)	BMI (kg/m2)	Surgical technique	Follow-up	Quality score (NOS)	Key finding
Ge et al., 2023[11]	Retrospective comparative (same-patient)	UKA versus HTO	15/15	59.6±2.9/60.3±4.2	2:13/2:13	26.9±3.6/27.4±10.6	Oxford UKA/medial open-wedge HTO	24 months	9	UKA higher FJS at 3–6 months; no difference ≥12 months
Huang et al., 2023[9]	Retrospective cohort	UKA versus HTO	128/111	63.3±3.7/61.7±5.7	19:109/20:91	26.3±3.4/25.4±3.3	Oxford UKA/medial open-wedge HTO	47 months	9	UKA showed higher mid-term FJS; age and FS predictive
Watanabe et al., 2022[10]	Multicenter retrospective	UKA versus HTO	48/48	73.8±5.2/61.3±9.8	9:39/12:36	24.1±2.8/26.1±3.8	Various UKA/medial open-wedge HTO	22 months	6	No significant difference in FJS
Jin et al., 2021[8]	Propensity- matched retrospective	UKA versus HTO	67/67	63.1±4.9/64.1±4.0	2:65/0:67	25.5±2.8/25.5±3.1	Fixed- bearing UKA/Medial open-wedge HTO	11.5 years	8	Comparable long-term FJS (UKA≈82; HTO≈78)
Maxwell et al., 2017[12]	Retrospective comparative	UKA versus HTO	95/75	<55 years (male only)	All male	NR	Cementless Oxford UKA/medial open-wedge HTO	6–8 years	6	Median FJS favored UKA (67 vs. 21)

UKA: Unicompartmental knee arthroplasty, HTO: High tibial osteotomy, NOS: Newcastle–Ottawa scale, FJS: Forgotten joint score-12, FS: Functional score, NR: Not reported

From the pooled demographic data, patients undergoing HTO were consistently younger than those undergoing UKA.^[8-12] A trend toward male predominance was observed in the HTO group, whereas a higher proportion of female patients underwent UKA. Body mass index (BMI) was comparable between groups, although several studies reported slightly higher BMI values in the HTO cohort. These differences likely reflect established selection criteria and indications for each procedure rather than study-related bias [Table 1].

Comparison of FJS at different time points (HTO vs. UKA)

UKA demonstrated higher FJS-12 than HTO during the early post-operative period (3–6 months), indicating lower joint awareness and faster functional recovery of the knee.^[11] By 12 months, this difference had narrowed substantially, reflecting progressive improvement in joint awareness following osteotomy [Table 2].^[11] At approximately 2 years of follow-up, FJS-12 values between UKA and HTO converged, and pooled analysis of three studies (Ge et al., 2023; Watanabe et al., 2022; Huang et al., 2023)^[9-11] demonstrated no statistically significant difference between the two procedures (MD: 2.39, 95% CI: −3.80–8.59; P = 0.45), despite increased heterogeneity across studies [Figure 2].

Table 2: FJS-12 score in UKA and HTO patients at different time points.

`Study	Procedure	Sample size	3 months	6 months	12 months	~24 months/mid-term
Ge et al., 2023[11]	UKA	15	51.4±8.0	69.2±6.0	69.4±6.0	69.7±6.0
Ge et al., 2023[11]	HTO	15	38.1±6.0	59.2±5.0	66.1±5.0	68.3±5.0
Huang et al., 2023[9]	UKA	128	-	-	-	66.7±7.4
Huang et al., 2023[9]	HTO	111	-	-	-	59.4±7.3
Watanabe et al., 2022[10]	UKA	48	-	-	-	60.0±23.9
Watanabe et al., 2022[10]	HTO	48	-	-	-	66.0±25.0
Maxwell et al., 2017*[12]	UKA		-	-	-	Median 67
Maxwell et al., 2017*[12]	HTO		-	-	-	Median 21
Jin et al., 2021[8]	UKA					81.9±15.1 (mean 11.9 years)
Jin et al., 2021[8]	HTO					77.9±17.0 (mean 11.1 years)

UKA: Unicompartmental knee arthroplasty, HTO: High tibial osteotomy, FJS: Forgotten joint score-12, *: Published abstract

Close

Figure 2:

Forest plot showing forgotten joint score-12 score among unicompartmental knee arthroplasty and high tibial osteotomy patients at 24 months. UKA: Unicompartmental knee arthroplasty, IV: Inverse Variance, HTO: High tibial osteotomy, SD: Standard deviation, CI: Confidence interval.

Export to PPT

Predictors of higher FJS-12 after HTO and UKA included younger age, lower body mass index, better post-operative pain relief, higher functional outcome scores (such as Knee Society Score, Functional Score, Knee Injury and Osteoarthritis Outcome Score, and Lysholm scores), and increased post-operative range of motion. Procedure type (HTO vs. UKA) was not a consistent predictor of long-term FJS-12, with its influence mainly limited to the early postoperative period [Table 3].

Table 3: Predictors of forgotten joint score (FJS-12) following UKA and HTO.

Study	Procedure	Positive predictors of higher FJS	Negative/non-significant predictors	Key notes
Ge et al., 2023[11]	UKA and HTO	Higher post-operative FS; increasing time from surgery	Age, BMI, ROM	Early UKA advantage; HTO improves over time
Huang et al., 2023[9]	UKA	Younger age; higher FS	BMI, sex, ASA, KL grade	UKA showed higher mid-term FJS
Huang et al., 2023[9]	HTO	Younger age; higher FS	BMI, sex, ASA, KL grade	Procedure type less influential
Watanabe et al., 2022[10]	UKA and HTO	Lower BMI; OA KL ≥3; osteonecrosis	Age, sex, ROM	No difference between procedures
Jin et al., 2021[8]	UKA and HTO	Balanced alignment; absence of advanced PF OA	—	Comparable long-term FJS
Maxwell et al., 2017[12]	UKA	Procedure type (UKA)	—	Young male cohort; median FJS
Maxwell et al., 2017[12]	HTO	—	High reoperation rate	Very low median FJS

BMI: Body mass index, UKA: Unicompartmental knee arthroplasty, HTO: High tibial osteotomy, NOS: Newcastle–Ottawa scale, FJS: Forgotten joint score-12, FS: Functional score, ROM: Range of motion, PF: Patellofemoral arthritis, KL: Kellgren-Lawrence, OA: Osteoarthritis

DISCUSSION

This systematic review and limited meta-analysis evaluated temporal trends in joint awareness following HTO and UKA using the FJS-12. The principal finding of this study is that although UKA is associated with significantly higher FJS values in the early post-operative period (3–6 months), these differences progressively diminish over time, and by approximately 2 years postoperatively, joint awareness is comparable between the two procedures. Even at long-term follow-up beyond 10 years, there was no difference in FJS score between the two procedures.

The FJS-12, originally developed and validated for joint arthroplasty, has been shown to be a valid, reliable, and sensitive patient-reported outcome measure in both UKA and HTO. In arthroplasty populations, including UKA, FJS-12 demonstrates excellent internal consistency and construct validity, with a substantially lower ceiling effect than conventional Patient-Reported Outcome Measures such as the Oxford knee score and Western Ontario and McMaster Universities Osteoarthritis Index, allowing superior discrimination among high-functioning patients.^[6,13] In the setting of medial opening-wedge HTO (MOWHTO), Itoh et al. validated FJS-12 in 71 patients and reported excellent internal consistency (Cronbach’s α = 0.9457), strong convergent validity with all KOOS subscales (r = 0.64–0.72), and absence of clinically relevant floor or ceiling effects (ceiling effect 8.5%).^[14] In contrast, significant ceiling effects were observed in KOOS pain, symptoms, and activities of daily living subscales.^[14] These findings confirm that FJS-12 reliably measures joint awareness not only after UKA but also following joint-preserving surgery such as HTO, where younger and more active patients frequently exceed the measurement capacity of traditional outcome scores.

This review demonstrated that patients undergoing HTO demonstrated lower FJS-12 in the early post-operative period compared with arthroplasty procedures, indicating greater joint awareness during the initial months (3–6 months) following surgery. This phenomenon is likely multifactorial. Unlike UKA, HTO is an extra-articular, bone-cutting procedure that necessitates osteotomy healing, prolonged protected weight-bearing, and gradual rehabilitation, all of which may delay restoration of normal proprioception and confidence in the limb. The presence of fixation hardware, local soft-tissue irritation around the medial tibia, and post-operative pain related to bone remodeling may further contribute to increased joint awareness.^[14] In the validation study of FJS-12 following MOWHTO, Itoh et al. highlighted the potential influence of fixation plates on joint perception.^[14] In their cohort, most patients underwent elective plate removal (100 out of 102) after osteotomy union, reflecting common Japanese orthopedic practice and lifestyle demands of Japanese patients involving frequent kneeling, deep knee flexion, and floor-based activities.^[14]The relatively bulky medial tibial plate was thought to irritate the overlying soft tissues during daily activities, which may partially explain the high acceptance of implant removal in this population. Such hardware-related factors may therefore play an important role in early joint awareness after HTO.

In addition, correction of limb alignment substantially alters knee biomechanics and load distribution, requiring neuromuscular adaptation that may take several months to normalize. These factors collectively contrast with UKA, where immediate joint resurfacing, preserved alignment, and earlier full weight bearing allow faster recovery of the knee into daily activities. Importantly, as osteotomy consolidation progresses and neuromuscular adaptation occurs, joint awareness after HTO improves steadily, explaining why differences in FJS-12 between HTO and arthroplasty tend to diminish or disappear by 1–2 years postoperatively.^[8-12,19]Notably, even at long-term follow-up beyond 10 years, Jin et al. reported no statistically significant difference in FJS-12 between UKA (mean FJS ≈82) and HTO (mean FJS ≈78).^[8]

The authors (Jin et al.) attributed the numerically higher, though statistically insignificant, FJS-12 following UKA to biomechanical and degenerative factors associated with HTO.^[8] Previous meta-analyses by Fu et al. have demonstrated that patients undergoing HTO often achieve greater valgus alignment compared with those undergoing UKA,^[20] a finding that was also observed in the long-term follow-up alignment data of Jin et al.^[8] Moreover, progression of patellofemoral (PF) joint osteoarthritis has been more frequently reported after HTO, particularly with MOWHTO, than after UKA. Several studies have shown that altered tibial slope and increased valgus alignment following HTO may increase PF contact pressures, potentially accelerating PF joint degeneration. However, the relationship between PF osteoarthritis progression and clinical outcomes such as function and anterior knee pain remains inconsistent across studies.^[21-27]

Comparative literature suggests that PF joint symptoms and degenerative changes tend to be less pronounced following UKA than HTO, and knee function appears to exert less influence on PF joint degeneration in UKA patients.^[28,29]While lateral tibiofemoral compartment degeneration has been less extensively studied, available evidence indicates a trend toward greater progression following HTO compared with UKA, although findings remain heterogeneous.^[28,29]Despite these biomechanical and radiographic differences, joint awareness at long-term follow-up of approximately 10 years remains statistically comparable between the two groups.^[8] Nevertheless, greater joint awareness might be anticipated in the HTO group over time (>10 years) owing to the higher likelihood of progression of PF and lateral compartment osteoarthritis.

Clinically, these findings have important implications for shared decision-making. While UKA may be favored in patients seeking faster early recovery and earlier joint forgetfulness, HTO remains a valid joint-preserving option that achieves equivalent joint awareness at mid-term follow-up in appropriately selected patients. Surgeons should therefore counsel patients that early post-operative joint awareness after HTO is expected and largely transient, rather than indicative of inferior long-term outcomes.

This study has several limitations. All included studies were retrospective, which introduces inherent selection bias and limits causal inference. There was heterogeneity in patient demographics, surgical techniques, rehabilitation protocols, and follow-up duration, which may have influenced FJS-12 outcomes. Differences in baseline characteristics, particularly age and activity level, reflect established selection criteria for HTO and UKA and may confound comparisons despite statistical adjustment. In addition, the number of studies reporting FJS-12 at uniform early post-operative time points was limited, restricting robust temporal analysis. Finally, long-term outcomes beyond mid-term follow-up remain insufficiently reported. Despite these limitations, this review is the first to introduce the concept of progression of joint forgetfulness in patients undergoing HTO and UKA during the post-operative period.

CONCLUSION

This systematic review and meta-analysis indicate that joint awareness following HTO and UKA is comparable at 2 years postoperatively. While early post-operative data suggest greater joint awareness after HTO at 3–6 months, these differences resolve by 1 year, resulting in equivalent FJS-12 at mid-term follow-up (>2 years). These findings suggest that, despite slower early functional integration, HTO achieves joint awareness outcomes comparable to UKA over time, supporting its continued role as a joint-preserving option in appropriately selected patients. Further prospective studies are warranted to confirm these temporal trends.