Robot-assisted mechanically aligned total knee arthroplasty: comparing early outcomes between varus and non-varus knees, and between CPAK Type I and Type IV subtypes

Abstract

Objective

This study aims to compare early pain relief and functional recovery between varus and non-varus phenotypes following robot-assisted mechanically aligned total knee arthroplasty (MA-TKA), with a specific focus on the two most common varus subtypes: Type I and Type IV.

Methods

A retrospective analysis was performed on 124 patients who underwent primary unilateral robot-assisted MA-TKA. Patients were classified according to the Coronal Plane Alignment of the Knee (CPAK) classification based on preoperative arithmetic hip-knee-ankle angle (aHKA) and joint line obliquity (JLO). Based on aHKA, patients were categorized into a varus group (CPAK Types I, IV, and VII, n = 93) and a non-varus group (neutral and valgus phenotypes, n = 31). Within the varus cohort, the two most prevalent subtypes — Type I (n = 64) and Type IV (n = 27) — were compared; Type VII (n = 2) was excluded from subtype-specific analyses due to small sample size. Patient-reported outcomes (VAS pain, KSS clinical/function, WOMAC) were evaluated preoperatively and at 1, 3, and 6 months postoperatively. Radiographic parameters (aHKA, MPTA, LDFA, JLO) were measured preoperatively and at 1 month.

Results

Preoperatively, the varus group had a significantly lower aHKA than the non-varus group (− 6.9° ± 2.4° vs. −0.1° ± 1.1°, p < 0.001). Postoperatively, both groups achieved neutral mechanical alignment with no significant difference in aHKA (p = 0.231). However, residual joint line obliquity differed significantly between Type I and Type IV knees (176.2° ± 0.2° vs. 179.3° ± 0.3°, p < 0.001). Non-varus patients had significantly better functional recovery at 1 month than varus patients (KSS‑Function: 60.5 ± 2.8 vs. 53.8 ± 3.1, p = 0.042; WOMAC: 36.2 ± 1.9 vs. 43.0 ± 2.3, p < 0.001), and these differences attenuated by 6 months. Within the varus subgroup, Type IV patients had lower 1‑month VAS scores and higher KSS‑Function scores than Type I patients (VAS: 3.5 ± 0.4 vs. 4.2 ± 0.4, p = 0.018; KSS‑Function: 55.0 ± 2.5 vs. 53.3 ± 2.9, p = 0.039).

Conclusions

Preoperative CPAK phenotype is associated with early recovery after robot-assisted MA‑TKA. Non-varus patients had higher 1‑month KSS‑Function scores than varus patients. Among varus knees, Type IV patients had lower 1‑month VAS scores and higher 1‑month KSS‑Function scores than Type I patients, whereas Type I knees retained greater residual postoperative joint line obliquity. Distinguishing varus subtypes may help refine prognostic expectations and guide individualized alignment strategies.

Introduction

Total Knee Arthroplasty (TKA) is currently internationally recognized as the most effective surgical intervention for end-stage knee osteoarthritis (OA), providing significant pain relief, deformity correction, and functional improvement. However, despite continuous advances in surgical techniques and implant design, patients’ subjective satisfaction with surgical outcomes has not reached an ideal level. Multiple large-scale studies and systematic reviews indicate that dissatisfaction rates after TKA remain as high as 10%--20%, with the proportion of patients experiencing residual pain ranging from 5.5% to over 50% [1, 2]. Common causes of dissatisfaction include residual postoperative pain, functional limitations, joint stiffness, and unmet preoperative expectations [3]. This clinical reality suggests that merely pursuing long-term implant survival does not equate to a favorable postoperative experience, and optimizing TKA clinical outcomes remains an important topic in the field of joint arthroplasty.

The traditional mechanical alignment (MA) strategy is currently the most widely adopted alignment philosophy in TKA [4]. Its core objective is to position the femoral and tibial components as perpendicular as possible to their respective mechanical axes, thereby restoring the overall lower limb alignment (hip-knee-ankle angle, HKA) to a neutral position of 0° ± 3°. This strategy has been validated over decades of clinical practice and is strongly associated with favorable long-term implant survival. However, Bellemans et al. first introduced the concept of “constitutional varus,” noting that a substantial proportion of the normal population naturally exhibits a mild varus alignment, and restoring a strictly neutral mechanical axis may not be the universally optimal choice [5]. In recent years, increasing evidence suggests that for patients with significant constitutional differences preoperatively, forcibly correcting to a neutral alignment using a uniform MA strategy may alter the native biomechanical environment of the knee and may not necessarily yield the best clinical outcomes [6, 7].

The Coronal Plane Alignment of the Knee (CPAK) classification was proposed by MacDessi et al. in 2021 [8]. Based on two parameters—arithmetic hip-knee-ankle angle (aHKA) and joint line obliquity (JLO)---this classification divides knees into nine distinct phenotypes, comprehensively reflecting the natural variation in limb alignment direction and joint surface obliquity across the population. The unique value of the CPAK classification lies in its ability not only to distinguish between varus and valgus directions but also to precisely differentiate the inclination status of the joint line. For instance, varus knees can be further categorized into those with a varus-inclined joint line (CPAK Type I, aHKA varus + JLO varus inclination) and those with a neutral or valgus-inclined joint line (CPAK Type IV, aHKA varus + JLO valgus inclination). Epidemiological data demonstrate that CPAK Types I and IV are the most common phenotypes in knee osteoarthritis patients, together accounting for more than 65%, suggesting that these phenotypes are of significant clinical research interest [9, 10].

Different CPAK phenotypes represent inherent constitutional characteristics of the knee. However, when TKA is performed using a uniform MA strategy, different knee phenotypes are subjected to the same alignment correction target (i.e., restoring aHKA to 0°), and the postoperative coronal plane phenotype will inevitably change. Previous studies have confirmed that long-term implant survival after TKA does not differ significantly based on preoperative phenotypic differences or postoperative phenotypic changes [11]. However, in the early postoperative period, differences in clinical outcomes resulting from preoperative phenotypic variation—particularly differences in the pace of pain relief and the trajectory of functional recovery—do objectively exist [11, 12]. Currently, systematic research and high-quality reports on this clinical phenomenon remain relatively scarce. Therefore, exploring the early recovery patterns of different CPAK phenotypes under a uniform MA strategy carries clear research significance.

Robot-assisted TKA technology is increasingly popularized in clinical practice due to its ability to accurately achieve the preoperatively planned alignment targets [11]. Compared with conventional manual TKA, robot-assisted technology can improve the reproducibility and accuracy of component positioning. This technical advantage largely eliminates confounding errors introduced by individual technical variability in manual procedures, providing an ideal technical platform for conducting homogenous clinical research [13]. Thus, by executing a uniform MA strategy through the highly standardized operative system of robot-assisted TKA, it becomes possible to more purely evaluate whether early postoperative clinical outcomes differ according to preoperative CPAK phenotype.

Based on the above background, this study aimed to: (1) compare patient-reported outcomes (PROMs), including pain (VAS), knee-specific function (KSS), and disease-specific health status (WOMAC), between varus and non-varus CPAK phenotypes at 1, 3, and 6 months postoperatively following robot-assisted MA-TKA; and (2) within the varus cohort, analyze the differences in recovery trajectories up to 6 months between CPAK Type I and Type IV, the two most prevalent varus subtypes.

Materials and methods

Study design and patient selection

This study was a single-center retrospective cohort study approved by the hospital ethics committee. A total of 135 consecutive patients who underwent primary unilateral robot-assisted total knee arthroplasty (TKA) for end-stage knee osteoarthritis between January 2024 and July 2025 were retrospectively collected. All surgeries were performed using a robotic-assisted system. The inclusion criteria were as follows: (1) confirmed diagnosis of primary knee osteoarthritis; (2) age greater than 55 years and less than 85 years; (3) availability of preoperative full-length standing hip-to-ankle radiographs suitable for CPAK classification; (4) completion of at least 6 months of postoperative follow-up with complete clinical data. The exclusion criteria included: (1) inflammatory arthritis; (2) post-traumatic arthritis; (3) prior history of high tibial osteotomy or distal femoral osteotomy; (4) presence of extra-articular deformity requiring concurrent corrective osteotomy; (5) incomplete radiographic or clinical follow-up data; (6) secondary knee arthropathy (including but not limited to inflammatory arthritis, post-traumatic arthritis, and arthropathy secondary to neoplastic or tumor-like diseases).

Based on the above inclusion and exclusion criteria, a total of 124 patients were ultimately included in the analysis. The patient screening and CPAK classification process is illustrated in Fig. 1.

Fig. 1.

Patient screening and CPAK classification flowchart. The diagram depicts the complete patient selection and grouping process for this retrospective study of robot-assisted primary total knee arthroplasty (TKA)

Based on the arithmetic hip-knee-ankle angle (aHKA) component of the CPAK classification, knees with varus alignment (aHKA < − 2°) were categorized into the varus group (CPAK Types I, IV, and VII) [8, 9]. Since Types I and IV are the two most prevalent varus phenotypes in osteoarthritic knees, these subtypes constituted the primary subgroups for comparative analysis [9, 10]. The remaining knees with neutral or valgus aHKA (CPAK Types II, III, V, VI, and IX) were combined as the non-varus reference group, as they collectively represent knees without preoperative varus deformity [12]. Type VII knees (n = 2) were included in the varus group for the primary varus-versus-non-varus comparison but were excluded from subtype-specific analyses due to the small sample size, consistent with previous CPAK studies [11].

Surgical technique and alignment strategy

All surgeries were performed by an experienced team of joint arthroplasty surgeons using a standard medial parapatellar approach under pneumatic tourniquet control. A posterior-stabilized (PS) cemented knee prosthesis system was uniformly used in all cases. The mechanical alignment strategy was strictly followed during robot-assisted TKA, with the specific planning as follows: the femoral component was positioned perpendicular to the femoral mechanical axis (target lateral distal femoral angle [LDFA] of 90°), and the tibial component was positioned perpendicular to the tibial mechanical axis (target medial proximal tibial angle [MPTA] of 90°). The overall lower limb alignment target was set to achieve an arithmetic hip-knee-ankle angle (aHKA) within the range of 0° ± 3°. Only the minimum necessary soft tissue release was performed intraoperatively to ensure satisfactory flexion and extension gap balance, and excessive release was avoided.

Radiographic assessment and CPAK classification

Weight-bearing full-length hip-to-ankle anteroposterior radiographs were obtained preoperatively and at 1 month postoperatively. Two independent observers performed all measurements using the hospital’s digital radiographic software. The intra- and inter-observer intraclass correlation coefficients (ICCs) for radiographic measurements ranged from 0.92 to 0.98, indicating excellent reliability. The measured parameters included the arithmetic hip-knee-ankle angle (aHKA), calculated as medial proximal tibial angle (MPTA) minus lateral distal femoral angle (LDFA); MPTA, the medial angle between the tibial mechanical axis and the tibial joint line; LDFA, the lateral angle between the femoral mechanical axis and the femoral joint line; and joint line obliquity (JLO), the angle between the femoral and tibial joint lines. Based on preoperative aHKA (varus: < −2°; neutral: −2° to 2°; valgus: 2°) and JLO (apex distal: < 177°; neutral: 177° to 183°; apex proximal: > 183°), all knees were classified into one of nine CPAK types according to the algorithm of MacDessi et al. [8]

Clinical outcome measures

Patient-reported outcome measures (PROMs) were prospectively collected at four time points: preoperatively and at 1, 3, and 6 months postoperatively. Pain intensity was assessed using a 100-mm Visual Analog Scale (VAS), with scores ranging from 0 (no pain) to 10 (worst imaginable pain) (Price et al., 1983) [14]. Knee-specific function was evaluated using the 2011 Knee Society Score (2011KSS), which comprises an objective knee score completed by the surgeon (range 0–100, higher scores indicating better knee function) and a functional activity score completed by the patient (range 0–100, higher scores indicating better function) (Scuderi et al., 2012) [15]. Overall disease-specific health status was measured using the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) total score (range 0–96, lower scores indicating better outcomes) (Bellamy et al., 1988) [16].

Statistical analysis

All statistical analyses were performed using IBM SPSS Statistics version 26.0 software. Continuous variables were expressed as mean ± standard deviation, and categorical variables were expressed as frequencies and proportions. For baseline data, continuous variables were compared using the independent samples t-test, and categorical variables were compared using the Pearson χ² test. For longitudinal changes in VAS, KSS, and WOMAC scores, two-way repeated measures analysis of variance (ANOVA) was performed, with group (varus/non-varus) or subgroup (Type I/Type IV) as the between-subject factor and time points (preoperative, 1 month, 3 months, and 6 months postoperatively) as the within-subject factor, to test for interaction effects. If significant, post hoc comparisons with Bonferroni correction were conducted. For radiographic parameters measured by two independent observers, the intraclass correlation coefficient (ICC) was used to assess intraobserver and interobserver reliability. All tests were two-tailed, and the significance level was set at p < 0.05.

Results

Patient characteristics and baseline data

After eligibility screening, 124 patients were included in the final analysis. The varus group comprised 93 knees (CPAK Type I, n = 64; Type IV, n = 27; Type Ⅶ, n = 2), and the non-varus group comprised 31 knees. Baseline demographic and preoperative clinical characteristics are summarized in Table 1. No significant differences were detected between the varus and non-varus groups, nor between the CPAK Type I and Type IV subgroups, in age, sex distribution, BMI, VAS pain score, KSS knee score, KSS function score, or WOMAC total score (all p > 0.05), confirming baseline comparability.

Table 1.

Baseline demographic characteristics and preoperative clinical data

Variable	Varus group (n = 93)	Non-varus group (n = 31)	P value	Type I (n = 64)	Type IV (n = 27)	P value
Age (years)	68.2 ± 7.1	67.5 ± 6.8	0.624	68.5 ± 7.3	67.6 ± 6.5	0.582
Sex (female/male)	62 / 31	20 / 11	0.833	42 / 22	18 / 9	0.921
Preoperative VAS	6.1 ± 1.0	6.0 ± 0.6	0.564	6.2 ± 1.0	6.0 ± 0.8	0.371
Preoperative KSS-Knee	38.7 ± 10.6	39.1 ± 9.4	0.852	38.1 ± 10.1	38.9 ± 11.3	0.741
Preoperative KSS-Function	31.5 ± 15.4	33.2 ± 12.8	0.591	30.4 ± 15.0	34.1 ± 15.9	0.295
Preoperative WOMAC	47.2 ± 13.2	49.8 ± 8.2	0.312	46.8 ± 14.1	48.9 ± 10.5	0.483

Data are presented as mean ± standard deviation or number of patients

VAS Visual Analog Scale for pain (0–10), KSS Knee Society Score (clinical and functional subscales, each 0–100), WOMAC Western Ontario and McMaster Universities Osteoarthritis Index (0–96)

P values were calculated using the independent samples t-test for continuous variables and the Pearson χ² test for sex distribution

Radiographic outcomes

Preoperative and 1-month postoperative radiographic parameters are presented in Table 2. Preoperative aHKA was significantly lower (more varus) in the varus group than in the non-varus group (− 6.9° ± 2.4° vs. −0.1° ± 1.1°, p < 0.001), accompanied by a smaller MPTA and larger LDFA (both p < 0.001). Following robot-assisted MA-TKA, both groups achieved the planned neutral mechanical alignment, with no significant intergroup differences in postoperative aHKA, MPTA, or LDFA. Preoperative JLO was comparable between groups (p = 0.988); however, at 1 month postoperatively, the non-varus group exhibited a significantly greater JLO than the varus group (179.6° ± 0.6° vs. 176.8° ± 1.1°, p < 0.001), indicating a less varus (or more valgus) joint line orientation. Within the varus cohort, CPAK Type IV knees had a substantially higher preoperative JLO (180.2° ± 0.6° vs. 175.3° ± 0.2°, p < 0.001), a constitutional difference that persisted after surgery (179.3° ± 0.3° vs. 176.2° ± 0.2°, p < 0.001).

Table 2.

Comparison of preoperative and 1-month postoperative radiographic parameters

Parameter	Time point	Varus (n = 93)	Non-varus (n = 31)	P value	Type I (n = 64)	Type IV (n = 27)	P value
aHKA (°)	Preoperative	−6.9 ± 2.4	−0.1 ± 1.1	< 0.001	−6.8 ± 2.5	−6.0 ± 2.0	0.142
	1 month postop	0.1 ± 0.7	0.3 ± 0.6	0.231	0.1 ± 0.8	0.1 ± 0.6	0.981
MPTA (°)	Preoperative	84.7 ± 1.9	88.4 ± 1.3	< 0.001	84.0 ± 1.4	87.2 ± 1.2	< 0.001
	1 month postop	89.5 ± 0.8	89.7 ± 0.4	0.198	89.5 ± 0.8	89.5 ± 0.8	0.982
LDFA (°)	Preoperative	91.9 ± 1.4	87.4 ± 0.7	< 0.001	91.3 ± 1.0	93.3 ± 0.9	< 0.001
	1 month postop	89.5 ± 0.8	89.7 ± 0.5	0.211	89.4 ± 0.9	89.7 ± 0.7	0.112
JLO (°)	Preoperative	175.5 ± 0.5	175.5 ± 1.8	0.988	175.3 ± 0.2	180.2 ± 0.6	< 0.001
	1 month postop	176.8 ± 1.1	179.6 ± 0.6	< 0.001	176.2 ± 0.2	179.3 ± 0.3	< 0.001

Data are presented as mean ± standard deviation

aHKA arithmetic hip-knee-ankle angle (negative values indicate varus), MPTA Medial proximal tibial angle, LDFA Lateral distal femoral angle, JLO Joint line obliquity (values < 177° indicate varus inclination, > 183° indicate valgus inclination), Postop postoperative

P values in bold denote statistical significance

Clinical outcomes: varus versus non-varus groups

Longitudinal patient-reported outcome scores are detailed in Table 3. Repeated‑measures ANOVA revealed significant group × time interaction effects for VAS pain, KSS function, and WOMAC scores (all p < 0.05), demonstrating divergent recovery trajectories (Fig. 2). Post hoc analyses showed that the non-varus group had a significantly lower VAS score (3.1 ± 0.2 vs. 4.0 ± 0.5, p = 0.031), higher KSS function score (60.5 ± 2.8 vs. 53.8 ± 3.1, p = 0.042), and lower WOMAC score (36.2 ± 1.5 vs. 43.0 ± 2.3, p < 0.001) at 1 month postoperatively. These between-group differences attenuated over time; at 6 months, VAS pain scores were comparable (p = 0.089), but the non-varus group still demonstrated statistically superior KSS function (p = 0.041) and lower WOMAC scores (p = 0.025).

Table 3.

Longitudinal clinical outcomes: varus versus non-varus groups

Time point	Outcome measure	Varus (n = 93)	Non-varus (n = 31)	P value
Postoperative 1 month	VAS	4.0 ± 0.5	3.1 ± 0.2	0.031
	KSS-Knee	67.0 ± 2.5	74.2 ± 1.5	0.011
	KSS-Function	53.8 ± 3.1	60.5 ± 2.8	0.042
	WOMAC	43.0 ± 2.3	36.2 ± 1.9	< 0.001
Postoperative 3 months	VAS	2.5 ± 0.3	2.5 ± 0.2	0.891
	KSS-Knee	77.5 ± 2.4	84.0 ± 1.0	< 0.001
	KSS-Function	67.9 ± 1.9	76.9 ± 1.6	< 0.001
	WOMAC	31.9 ± 1.9	26.1 ± 1.5	0.007
Postoperative 6 months	VAS	1.3 ± 0.3	1.0 ± 0.1	0.089
	KSS-Knee	86.1 ± 1.7	87.2 ± 1.5	0.123
	KSS-Function	79.3 ± 1.2	82.0 ± 1.6	0.041
	WOMAC	22.7 ± 1.2	19.2 ± 1.1	< 0.001

Data are presented as mean ± standard deviation

VAS Visual Analog Scale for pain, KSS Knee Society Score, WOMAC Western Ontario and McMaster Universities Osteoarthritis Index 

Group × time interactions were significant for VAS, KSS-Function, and WOMAC (all p < 0.05). Post hoc between-group comparisons at each time point were performed with Bonferroni correction

Fig. 2.

Group-by-time interaction analysis of clinical outcome measures following robot-assisted primary total knee arthroplasty. Effects of group-by-time interaction on clinical outcome measures in patients undergoing robot-assisted primary total knee arthroplasty . Abbreviations for each time point are as follows: pre-op: preoperative; 1m、3m and 6m refer to 1 month, 3 months, and 6 months postoperatively, respectively

Subgroup analysis: CPAK Type I versus Type IV

Within the varus group, Type IV knees exhibited a more favorable early recovery compared with Type I knees (Table 4). Significant group × time interactions were observed for VAS and KSS function (p < 0.05). At 1 month, Type IV patients had a significantly lower VAS score (3.5 ± 0.4 vs. 4.2 ± 0.4, p = 0.018) and higher KSS function score (55.0 ± 2.5 vs. 53.3 ± 2.9, p = 0.039). The VAS advantage for Type IV persisted at 3 months (p = 0.035), but no significant between-subtype differences were detected at 6 months for any outcome measure. WOMAC scores did not differ significantly between the two subtypes at any time point.

Table 4.

Subgroup analysis: CPAK Type I versus Type IV

Time point	Outcome measure	Type I (n = 64)	Type IV (n = 27)	P value
Postoperative 1 month	VAS	4.2 ± 0.4	3.5 ± 0.4	0.018
	KSS-Knee	67.2 ± 2.4	67.9 ± 2.1	0.213
	KSS-Function	53.3 ± 2.9	55.0 ± 2.5	0.039
	WOMAC	43.4 ± 2.0	41.5 ± 2.1	0.185
Postoperative 3 months	VAS	2.6 ± 0.3	2.2 ± 0.2	0.035
	KSS-Knee	77.8 ± 2.2	78.3 ± 2.0	0.342
	KSS-Function	67.6 ± 1.8	68.7 ± 1.7	0.287
	WOMAC	32.1 ± 1.8	30.8 ± 1.9	0.225
Postoperative 6 months	VAS	1.3 ± 0.3	1.2 ± 0.3	0.412
	KSS-Knee	86.0 ± 1.6	86.5 ± 1.4	0.156
	KSS-Function	79.3 ± 1.1	79.6 ± 1.3	0.612
	WOMAC	22.8 ± 1.1	22.1 ± 1.2	0.314

Data are presented as mean ± standard deviation

VAS Visual Analog Scale for pain, KSS Knee Society Score, WOMAC Western Ontario and McMaster Universities Osteoarthritis Index

Group × time interactions were significant for VAS and KSS-Function (p < 0.05). Post hoc between-subtype comparisons at each time point were performed with Bonferroni correction. P values in bold denote statistical significance

Discussion

This study demonstrates that non-varus CPAK patients had higher 1‑month KSS‑Function scores than varus patients following robot‑assisted MA‑TKA, although this between‑group difference had largely attenuated by 6 months. More importantly, within the varus subgroup, Type IV patients had lower 1‑month VAS scores and higher 1‑month KSS‑Function scores than Type I patients. These differences emerged against a background in which overall limb alignment had been corrected to approximately 0° in all groups, yet residual joint line obliquity diverged considerably. Given the limited evidence on this topic [11, 17], these hypothesis-generating findings suggest that preoperative CPAK phenotype is associated with the pace of early recovery, and that residual joint line obliquity may be a relevant factor when interpreting postoperative outcomes.

In this cohort, both the varus and non-varus groups achieved neutral overall limb alignment postoperatively; however, significant differences in joint line obliquity persisted between Type I and Type IV knees. The non-varus group demonstrated better early functional scores than the varus group, although this difference diminished by 6 months. These findings are consistent with existing literature reporting that a change in CPAK category after TKA is associated with inferior early patient-reported outcomes [18]. Within the varus group, the more favorable early recovery trajectory of Type IV relative to Type I echoes the observation by Sahbat Y et al. that preservation of native joint line orientation correlates with milder postoperative pain [19].

Prior finite element studies have demonstrated that joint line obliquity exceeding 3° increases medial compartment contact stress [20]. Clinical studies have further reported that postoperative joint line varus inclination is associated with worse pain scores and slower functional recovery, with potential mechanisms involving altered medial soft-tissue tension and patellofemoral maltracking [21, 22]. In MA-TKA for Type I knees, the deformity originates predominantly from the proximal tibia; correcting the tibial varus to achieve a neutral mechanical axis often requires extensive medial release [23, 24], while the femur exhibits only mild compensatory valgus, such that the joint line frequently remains varus-inclined after tibial resection and soft-tissue balancing. In contrast, the deformity in Type IV knees originates more from the femoral side, and adjusting the distal femoral cut can restore the joint line to near-neutral [25]. This difference in the site of bony correction may partly explain the divergent early clinical courses.

Moreover, satisfactory early pain relief and functional recovery are closely linked to balanced mediolateral soft-tissue tension and undisturbed ligament healing [26]. In MA-TKA, excessive medial release may prolong the inflammatory response and delay the restoration of dynamic stabilizers in the early postoperative period. Although robotic assistance improves osteotomy precision, the ability to alter the inherent joint line orientation remains constrained by the patient’s native articular geometry and the fixed resection planes determined by preoperative planning [27, 28]. A neutral overall aHKA can be masked by an unbalanced combination of MPTA and LDFA—for example, an MPTA of 92° compensated by an LDFA of 88°---an appearance that obscures residual joint line obliquity [29]. Such compensatory patterns are more likely to occur in Type I knees and may persistently generate asymmetric medial compartment loading during early weight bearing. Taken together, these observations suggest that after achieving the target mechanical alignment, the soft-tissue environment around residual joint line obliquity may modulate the tempo of symptomatic recovery.

Several limitations should be acknowledged. First, the retrospective, single-center design precludes causal inference; the observed associations require prospective validation. Second, the CPAK Type IV subgroup (n = 27) was relatively small, limiting the statistical power of subtype-specific comparisons and warranting cautious interpretation of these findings. Third, the 6-month follow-up period, while adequate for capturing early recovery trajectories, is insufficient to determine whether the observed differences persist long-term or affect implant survival. Fourth, the non-varus group combined neutral and valgus phenotypes, consistent with the aHKA-based categorization used in prior CPAK studies [8, 12, 27]; only two valgus knees were included, so heterogeneity was minimal. Fifth, perioperative covariates such as soft-tissue release extent, rehabilitation compliance, and psychological factors were not collected. Finally, the single-implant design limits generalizability to other implant types. These findings should therefore be considered hypothesis-generating rather than definitive.

Conclusions

In conclusion, preoperative CPAK subtyping is associated with different early recovery trajectories after robot-assisted MA-TKA, primarily driven by residual joint line obliquity. In this cohort, CPAK Type IV patients experienced less pain and better function at 1 month than Type I patients, suggesting that distinguishing these varus subtypes provides meaningful prognostic information for clinicians. While these findings do not yet mandate a shift from mechanical alignment, they underscore that achieving a neutral mechanical axis alone may not fully optimize early patient-reported outcomes; preoperative awareness of subtype-specific residual obliquity can help set realistic recovery expectations. Prospective studies with longer follow-up are required to validate these hypothesis-generating results before any change in clinical practice can be recommended.

Abbreviations

aHKA

Arithmetic hip-knee-ankle angle

ANOVA

Analysis of variance

BMI

Body mass index

CPAK

Coronal Plane Alignment of the Knee

HKA

Hip-knee-ankle angle

ICC

Intraclass correlation coefficient

JLO

Joint line obliquity

KSS

Knee Society Score

LDFA

Lateral distal femoral angle

MA

Mechanical alignment

MA-TKA

Mechanically aligned total knee arthroplasty

MPTA

Medial proximal tibial angle

OA

Osteoarthritis

PROMs

Patient-reported outcome measures

PS

Posterior-stabilized

TKA

Total knee arthroplasty

VAS

Visual Analog Scale

WOMAC

Western Ontario and McMaster Universities Osteoarthritis Index

Authors’ contributions

Pengpeng Li: study design and paper writing. Jihe Ban: data collection and paper writing. Yunhai Liu: data collection and paper writing. Juan Li: data analysis and paper writing. Jichao Hu: study design, data analysis and paper writing. All authors read and approved the final manuscript.

Funding

This study was supported by the Science and Technology Plan Project of Jiaxing (grant number: 2024AD30009).

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the Declaration of Helsinki. The retrospective study was approved by the Ethics Committee of Zhejiang Rongjun Hospital (Approval No. R2026-072), and the requirement for informed consent was waived due to the retrospective nature of the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.