Unicompartmental knee arthroplasty in patients with Parkinson’s disease

Abstract

Introduction

Patients with Parkinson’s disease (PD) are a vulnerable subgroup facing elevated risks of complications and functional decline following knee arthroplasty. However, data on the outcomes of minimally invasive unicompartmental knee arthroplasty (UKA) in this population are limited. This study´s purpose was to assess perioperative complications, implant revision-free and reoperation-free survivorship as well as functional outcome in PD patients following UKA.

Materials and methods

In this retrospective single-center study, 42 knees in 39 patients with PD who underwent medial or lateral UKA between 2016 and 2022 were analyzed. The mean age was 70.6 ± 9.1 years, and the mean BMI was 27.9 ± 5.2 kg/m². A total of 26 medial and 16 lateral UKAs were performed, with a minimum follow-up of two years (mean 5.0 ± 2.0). All medical complications were recorded. Implant survivorship (tibia and/or femur) and reoperation-free survival were evaluated using Kaplan-Meier analysis, allowing estimation of long-term survival beyond the mean follow-up duration. Functional outcomes were assessed using the Oxford Knee Score (OKS) and the UCLA Activity Score.

Results

Of the 42 knees, 6 (14.3%) underwent reoperation, including 4 cases (9.5%) requiring implant revision. The cumulative 9-year implant survivorship was 90.5% (95% CI: 81.7–99.3), and reoperation-free survival was 85.7% (95% CI: 75.1–96.3), respectively. No perioperative cardiovascular complications occurred. OKS improved significantly from 16.2 ± 5.5 to 39.6 ± 7.1 (p = 0.027), while the UCLA Activity Score showed a trend towards improvement from 4.0 ± 2.0 to 5.0 ± 1.7 (p = 0.078).

Conclusion

In this observational study, UKA in patients with PD was associated with favorable implant survivorship and encouraging functional outcomes. Considering the very low medical complication rate observed in this study, UKA may represent a viable treatment option for isolated end-stage unicompartmental osteoarthritis in carefully selected patients.

Introduction

Parkinson’s disease (PD) is the second most common neurodegenerative disorder, with a steadily rising prevalence in the aging population [1]. Characterized by a gradual onset of motor symptoms such as bradykinesia, rigidity, and postural instability, PD primarily affects individuals over the age of 50 and is more prevalent in men [2]. As life expectancy continues to rise, the incidence of end-stage knee osteoarthritis (OA) is expected to increase among patients with PD [1, 3].

Total knee arthroplasty (TKA) and unicompartmental knee arthroplasty (UKA) are both well-established surgical treatment options for end-stage knee OA. However, several studies have reported that patients with PD undergoing TKA are at increased risk for perioperative complications and mortality, along with less favorable outcomes compared to the general population [4–8]. In contrast, UKA represents a less invasive alternative for isolated unicompartmental disease, offering lower perioperative mortality and morbidity, faster recovery, better functional outcome and a more physiological gait compared to TKA [9–12]. These characteristics may be particularly advantageous in patients with Parkinson’s disease, who often have multiple medical comorbidities and reduced functional reserve. At the same time, Parkinson’s disease represents a heterogeneous condition with varying disease severity, which may influence the outcome of knee arthroplasty. Therefore, UKA appears as an attractive minimally invasive surgical treatment option for selected PD patients with an indication for knee arthroplasty. To date, only one small-cohort study has investigated the clinical results of UKA in patients with PD [13].

The present study therefore aims to analyze the outcome of UKA in patients with PD. We hypothesized that UKA in this patient population results in low perioperative complication rates, acceptable survivorship, and favorable functional outcomes.

Methods

Study Cohort

In this retrospective single-center study, all patients who underwent primary UKA (total cohort n = 6,743; medial n = 5,260; lateral n = 1,483) between January 2016 and December 2022 were screened for the presence of PD at the time of surgery. Parkinson’s disease was identified based on a documented clinical diagnosis in the medical records. A total of 54 consecutive knees in 50 patients with a minimum follow-up of two years were identified. Patients that had died before 2 years of follow-up were not included in this study. Of these, 45 patients were successfully contacted, resulting in a follow-up rate of 90%. Six patients with UKA in situ did not want to participate in the study and were excluded, leaving 42 knees in 39 patients for final analysis (Table 1). The study was approved by the institutional review board (IRB_003/2023), and all included patients consented to participate in the study.

Table 1.

The demographic data of all included patients and surgical details, are presented as numbers (percentage) or mean ± standard deviation (range)

Demographic Data (n = 39)
Men	21 (54)
Women	18 (46)
Age (yrs)	70.6 ± 9.1 (47–87)
BMI (kg/m2)	27.9 ± 5.2 (18.8–45.7)
Follow-up (yrs)	5.0 ± 2.0 (2.5–8.7)
ASA I	0
ASA II	16 (41)
ASA III	23 (59)
ASA IV	0
Parkinson Medication (n = 36)	30 (83) Levodopa-containing
	- 9 (30) Monotherapy
	- 21 (70) Combination (≥ 2 agents)
	6 (17) Non-Levodopa-containing
	− 4 (67) Dopamine agonist
	− 2 (33) MAO-B inhibitor
Surgical Details (n = 42)
Implant type	16 (38) lateral FB (Oxford) 20 (48) cemented medial MB (Oxford) 6 (14) cementless medial MB (Oxford)
Surgery time (min)	Medial: 46.7 ± 14.1 (26–72) Lateral: 46.2 ± 12.4 (28–75)
Length of stay (days)a	5 (3–23)

ASA American Society of Anesthesiologists, BMI body-mass-index, FB fixed-bearing, MAO-B Monoamine Oxidase B, MB mobile-bearing

^aStated as median value (range)

Indication and surgical technique

The standard radiographic workup for all knee arthroplasty patients included true AP/lateral knee radiographs, patella skyline views, varus/valgus stress views, and AP standing hip-to-ankle radiographs. The Oxford criteria were applied for the indication of medial UKA [14] and the same principles were applied for lateral UKA. In selected cases, when valgus stress views did not show lateral bone-on-bone wear, a Rosenberg view (45°) and/or MRI was obtained to confirm relevant full-thickness lateral-compartment cartilage defects. Overall limb alignment was not considered a relevant factor for the indication of UKA. UKA was offered to all patients with isolated end-stage compartmental OA, with a correctable intra-articular deformity, functionally intact anterior cruciate ligament and preserved cartilage in the contralateral compartment as confirmed by stress radiographs. Fixed flexion contractures of less than 15° were tolerated. Body weight, age, activity level, and the presence of chondrocalcinosis in the contralateral compartment were not considered contraindications for UKA [15]. The patellofemoral joint status was neglected unless patients presented with primary patellofemoral symptoms (lateral patellar bone loss in varus knees, medial patellar bone loss in valgus knees; grooving or maltracking). A total of 13 patients (31%; 13/42) had undergone prior knee surgery, including five partial meniscectomies, five unspecified knee arthroscopies, one anterior cruciate ligament reconstruction, and two open meniscectomies.

A minimally invasive medial or lateral parapatellar approach without dislocation of the patella was used in all cases. The mobile-bearing (MB) Oxford Partial or the cemented Oxford Fixed-Lateral Partial Knee (both Zimmer Biomet, Warsaw, Indiana, USA) were implanted. Out of 26 medial UKAs, 6 (23%) received a cementless MB implant. In cementless cases, a bone-preserving tibial cut was performed and the tibial component was placed in slight varus to reduce the risk of periprosthetic tibial fractures and valgus subsidence [16]. Postoperatively, early mobilization on the day of surgery with full weight-bearing as tolerated was encouraged. Rehabilitation focused on gradual range-of-motion exercises, muscle strengthening, and gait training under physiotherapeutic supervision. However, patients were advised to avoid any forced knee flexion and leg press exercises in the early rehabilitation after lateral UKA to protect the thin lateral joint capsule.

Functional and radiological outcome

Between February 2024 and May 2025, patients were contacted for a standardized telephone interview to determine implant survival, any reoperation on the same knee, as well as the postoperative Oxford Knee Score (OKS) [17, 18] and the University of California, Los Angeles (UCLA) Activity Score [19]. The patient acceptable symptom state (PASS) for OKS, defined as the threshold at which patients consider themselves well, has been reported to be around 31 [20]. The date of contact was defined as final follow-up time. Preoperative OKS and UCLA Activity Score were available in 6 (14%; 6/42) knees. Postoperative OKS and UCLA Activity Score were available in 32 (76%; 32/42) knees. Postoperative PROMs were missing in 9 patients (10 knees) that had passed away. However, information on implant survival was available in all 42 included knees. Data for medical complications was obtained from the institutional medical records.

Preoperatively, the mechanical HKAA and the MPTA were evaluated on AP standing hip-to-ankle radiographs. The HKAA was defined as the angle between the femoral mechanical axis (from the center of the hip to the center of the knee) and the tibial mechanical axis (from the center of the knee to the center of the ankle) [21]. MPTA was defined as the angle between the tibial mechanical axis and the best-fit line along the tibial plateau surface [21].

Postoperative hip-to-ankle radiographs were analyzed using OsiriX (Pixmeo, Geneva, Switzerland) with a validated and previously described reconstruction technique [22]. Intraobserver reliability of postoperative HKAA showed excellent agreement (intraclass correlation coefficient [ICC] 0.998, 95% CI 0.994–0.999), while interobserver reliability was similarly high (ICC 0.981, 95% CI 0.952–0.992). All postoperative radiographs were obtained within two days after index surgery and evaluation included mechanical HKAA and Joint Line Obliquity (JLO). JLO was measured between the tibial mechanical axis and a line connecting the most distal point of the femoral component and the midpoint of the lateral/medial joint space [23]. The preoperative MPTA in end-stage unicompartmental OA was considered comparable to the postoperative JLO.

Statistical analysis

Continuous variables and scores are presented as mean ± standard deviation (SD), while categorical variables are expressed as absolute numbers and percentages. The normality of the data was assessed using the Kolmogorov-Smirnov test. Depending on the distribution, either an independent t-test or a Mann-Whitney U-test was applied for unpaired comparisons. For paired data, a paired t-test or Wilcoxon signed-rank test was used accordingly. Kaplan-Meier survival analysis was employed to calculate cumulative survival rates at 9 years postoperatively, allowing estimation of long-term survival beyond the mean follow-up duration. The endpoints were (1) implant revision and (2) any reoperation. Implant revision was defined as the exchange of any tibial and/or femoral component. For the Kaplan-Meier survival analysis, patients were censored at death when the implant was in situ. Given the small proportion of bilateral cases (3 of 39 patients), no correction for within-patient dependency was performed. A p-value of less than 0.05 was considered statistically significant. All statistical analyses were performed using IBM SPSS Statistics for Windows, version 30.0.0.0 (IBM Corp., Armonk, NY, USA).

Results

Across all 42 UKAs, two (5%; 2/42) medical complications were observed: one patient experienced a postoperative epileptic seizure in the context of pre-existing epilepsy, and another patient had a postoperative recurrence of chronic erysipelas requiring intravenous antibiotics. No perioperative cardiovascular complications occurred. No patient required readmission during a 90-day period. The mean hemoglobin level decreased moderately from 14 g/dL preoperatively to 13 g/dL postoperatively. No patient required postoperative blood transfusion. At the time of surgery, 36 patients (92%; 36/39) were receiving antiparkinsonian medication. Of these, 30 (83%; 30/36) were treated with a Levodopa-containing regimen, while 6 (17%; 6/36) received non–Levodopa-based therapy (Table 1).

The survival analysis demonstrated a cumulative 9-year implant survivorship of 90.5% (95% CI: 81.7–99.3). The cumulative 9-year reoperation-free survivorship was 85.7% (95% CI: 75.1–96.3) (Fig. 1). A total of six reoperations (14.3%; 6/42) were observed within the first two postoperative years including four implant revisions (9.5%; 4/42) (Table 2).

Fig. 1.

Survival curve for (A) implant revision with 90.5% (95% CI: 81.7–99.3) and (B) reoperation with 85.7% (95% CI: 75.1–96.3) at 9 years postoperatively

Table 2.

Detailed information on the four implant revisions and two reoperations. For HKAA, negative values indicate valgus alignment, and positive values indicate varus alignment

	Implant Revision (n = 4)				Reoperation (n = 2)
	Case 1	Case 2	Case 3	Case 4	Case 5	Case 6
Implant Type	Lateral Oxford cemented	Medial Oxford cemented	Medial Oxford cementless	Medial Oxford cemented	Lateral Oxford cemented	Lateral Oxford cemented
Reason for revision	Medial and patellofemoral disease progression	Lateral disease progression	Atraumatic periprosthetic fracture	Aseptic tibial loosening	Medial disease progression	Early infection with Staphylococcus aureus
Treatment	Conversion to primary TKA	Conversion to TKA with medial augment/stem extension	Conversion to TKA with medial augment/tibial cone	Exchange of tibial implant (Vanguard M)	Additional medial UKA (Vanguard M)	Two-stage DAIR
Time to revision	19 months	15 months	2 weeks	8 months	4 months	2 weeks
Sex	female	female	male	female	male	female
Age	70	64	75	47	69	77
Postoperative OKS	43	37	n.d.	30	48	n.d.
Postoperative UCLA	7	4	n.d.	5	7	n.d.
Preoperative HKAA (°)	-7.9	10.5	9.5	5.1	-5.4	-12.4
Postoperative HKAA (°)	-3.5	2.4	3.9	-0.9	4.1	-2.9
BMI (kg/m2)	29.1	45.7	27.7	30.6	24.2	24.6

BMI Body-Mass-Index, DAIR Debridement, Antibiotics and Implant Retention, FB fixed-bearing, HKAA Hip-Knee-Ankle Angle, n.d. not determined, OKS Oxford Knee Score, TKA Total Knee Arthroplasty, UCLA University of California Los Angeles Activity Score, UKA Unicompartmental Knee Arthroplasty

OKS improved significantly from a preoperative mean of 16.2 ± 5.5 (range, 12–26) to a postoperative mean of 39.6 ± 7.1 (range, 23–48) (p = 0.027). The UCLA Activity Score showed a trend toward improvement from 4.0 ± 2.0 (range, 2–7) to 5.0 ± 1.7 (range, 2–7) (p = 0.078). For medial UKA, HKAA was corrected from a preoperative mean of 8.5° ± 2.7 varus (range, 4.1° to 15.5° varus) to a postoperative mean of 3.8° ± 2.8 varus (range, 0.9° valgus to 8.8° varus) (p < 0.001). The preoperative MPTA in varus knees averaged 84.9° ± 2.4 (range, 79.3° to 89.6°), while postoperative JLO was 86.2°± 2.6 (range, 81.3° to 90.3°) (p = 0.008). In lateral UKA, HKAA was corrected from 8.5° ± 3.3 valgus preoperatively (range, 12.4° to 2.2° valgus) to 2.9° ± 2.9 valgus postoperatively (range, 8.5° valgus to 4.1° varus) (p < 0.001). The preoperative MPTA was 89.7° ± 2.5 (range, 84.5° to 94.5°) and changed to 88.0° ± 1.6 (range, 85.1° to 91.1°) postoperatively (p = 0.023).

Discussion

This study highlights the clinical value of UKA as a surgical treatment option for end-stage unicompartmental knee OA in patients with Parkinson’s disease. The present study presents favorable implant survivorship, clinically relevant functional improvements, and a very low perioperative medical complication rate for UKA in this distinct patient population.

Patients with Parkinson’s disease are known to have an increased risk of perioperative medical complications following TKA. Large database studies have shown that PD patients undergoing TKA experience higher rates of postoperative delirium, pneumonia, urinary tract infection, need for blood transfusion, and increased 90-day readmission rates compared with non-PD controls [4, 24]. While previous reports have demonstrated lower medical complication rates after UKA compared with TKA in the general population, no separate analysis has been performed specifically for PD patients [9, 25]. In our series, the overall rate of medical complications after UKA was very low, and none of the commonly reported complications observed in PD patients after TKA occurred.

To date, there is only one study on the usage of medial UKA in 13 PD patients describing modest functional improvement and one revision due to disease progression at 10-year follow-up [13]. The present study expands on these findings based on a substantially larger cohort of medial and lateral UKA demonstrating an implant survivorship of 91% and a reoperation-free survivorship of 86% at 9 years. In contrast to Goh et al., this study also included knees with surgery on the same knee. While direct comparisons are limited by the absence of a control group, the observed survivorship estimates appear not inferior to previously reported outcomes of primary TKA in patients with Parkinson’s disease with implant survival rates between 66% and 90% [6–8] Interestingly, failure mechanisms following knee arthroplasty in patients with PD appear to differ from those observed in the general population. Recent studies reported higher rates for periprosthetic joint infection (PPI) as well as periprosthetic fracture (PPF) following arthroplasty procedures in patients with PD [8, 24].

In our cohort, one early postoperative PPI occurred following lateral UKA, corresponding to an infection rate of 2%. The infection rates after primary UKA in the general population have consistently been described around 0.3–0.9% [26, 27]. However, due to the limited number of events in the present study, no meaningful comparison can be made, and no conclusions regarding relative infection risk should be drawn. Gait disorder commonly observed in Parkinson’s disease have been associated with an increased incidence of falling, which may contribute to higher rates of periprosthetic fracture after knee arthroplasty [8, 24]. Recent studies have reported a higher incidence of PPF, particularly in the early postoperative period, for cementless medial UKA compared to cemented medial UKA. This severe complication is typically surgery-related and associated to female sex, age > 70 years, BMI > 40, and valgus alignment of the tibial component (reduced keel–cortex distance posteriorly) [16, 28]. In our series, one periprosthetic fracture in a 75-year-old male patient was detected 13 days after cementless medial UKA without any reported fall. In this patient, severe medial-compartment OA with bone loss of the proximal tibia, a deep vertical cut combined with valgus orientation of the tibial component relative to the anatomy of the proximal tibia may have predisposed to the fracture.

OA progression is considered the most common reason for revision after UKA [29, 30]. In line with the current literature, disease progression was also the most frequent cause (3/6 cases) for additional surgery in this study. Overcorrection is a well described risk factor for early femorotibial disease progression [31]. In this study, overcorrection to 4° varus in a valgus knee resulted in medial-compartment OA four months postoperatively (Table 2, Case 5). In another case, a high BMI of 46 kg/m² may have contributed to disease progression observed after 15 months (Table 2, Case 2) [32]. Due to the small numbers and the lack of a control group, this study cannot establish a causal link between PD and increased disease progression rates after UKA. The observed cases of disease progression (all within the first 2 postoperative years) appear rather surgery-related or must be interpreted in the context of a non-ideal indication. Our findings therefore do not suggest any reason to discourage the use of UKA in patients with PD, particularly given its potential advantages as a less invasive procedure.

Patients with Parkinson’s disease undergoing mobile-bearing UKA may theoretically face an increased risk of bearing dislocation in medial UKA due to tremor, impaired postural control and altered joint kinematics. Interestingly, none of these disease-specific neuromuscular impairments translated into a bearing dislocation. These findings therefore support the continued use of mobile-bearing implants in PD patients.

Achieving good functional results following knee arthroplasty in patients with PD remains a clinical challenge. Previous studies have reported inferior functional outcomes in PD patients compared to controls following TKA [5, 6]. Similarly, Goh et al. reported improvements in pain scores but only modest gains in knee function following UKA in a series of 13 PD patients, with relevant functional improvement observed in only half of the patients [13]. In contrast, the postoperative OKS scores in the present study were encouraging, showing a significant improvement to nearly 40 points at a mean follow-up of 5 years; 80% (26/32) of patients exceeded the PASS threshold, with only six patients falling below it. However, as Parkinson’s disease–related gait and neuromuscular impairments have been shown to worsen over time and to progress largely independent of dopaminergic medication [33], the long-term impact on functional outcomes after medial UKA remains uncertain. To optimize Parkinson’s disease control and enhance the functional outcome, we strongly encourage a multidisciplinary approach in close collaboration with a neurologist during surgical planning, perioperative management, and postoperative rehabilitation.

This study has several limitations that warrant consideration. First, its retrospective design may introduce selection and information bias. Second, although the study presents the largest cohort to date examining the outcome of UKA in PD patients, the sample size remains relatively small, which limits statistical power and the ability to perform meaningful subgroup analyses. Third, only 76% of patients completed postoperative PROMs, as a substantial proportion had passed away at the time of follow-up. In addition, preoperative PROMs were available only for a small subset of patients, as routine collection was implemented late in the study period, limiting paired comparisons. Fourth, the severity of PD symptoms (e.g., Hoehn and Yahr stage) was not systematically assessed, which may have influenced clinical outcomes and limits the interpretability of the results. Fifth, a selection bias cannot be excluded due to a lack of randomization for the treatment of UKA versus TKA. Lastly, Kaplan–Meier–estimated survivorship beyond the mean follow-up should be interpreted cautiously due to the declining number of patients at risk over time.

Conclusion

Unicompartmental knee arthroplasty in patients with Parkinson’s disease was associated with favorable implant survivorship and encouraging functional outcomes in this observational study. In the absence of a control group, UKA provides a low perioperative complication rate and is therefore a potentially reasonable treatment option for isolated end-stage unicompartmental OA in carefully selected patients. .

Author contributions

CS Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Visualization; Writing - original draft, TK Conceptualization; Methodology; Project administration; Software; Writing - review & editing, SA Conceptualization; Writing - review & editing, JH Conceptualization; Writing - review & editing, PA Conceptualization; Writing - review & editing, CM Conceptualization; Methodology; Project administration; Supervision; Validation; Writing - review & editing, WW Conceptualization; Methodology; Project administration; Supervision; Validation; Writing - review & editing.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Conflict of interest

PA & JH are paid consultants for Zimmer Biomet, and CM is a paid consultant for Medacta. PA & CM offer research support for Zimmer Biomet. All other authors declare no competing interests.