Mid-Flexion Instability in Total Knee Arthroplasty: Insights from Robotic-Assisted Surgery

Abstract

Despite technological advancements with robotic-assisted surgery, instability remains a challenge in total knee arthroplasty (TKA). Mid-flexion instability (MFI) has been reported to cause patient dissatisfaction. With no universal diagnostic criteria, the MFI concept is still ambiguous, and no specific treatment algorithm is defined. This study aims to analyze the MFI concept and risk factors and investigate how robotic surgery, compared to manual TKA, could impact the MFI concept. A comprehensive investigation of the current literature regarding MIF, focusing especially on its relationship with robotic surgery TKA, was conducted using the PubMed and Scopus databases. The MIF concept remains poorly understood, so it is crucial to prevent it by recognizing risk factors, which are technique-related, implant-related, and patient-related. Since robotics offers optimal balancing in TKA and reduces causes affecting MFI, it could indirectly impact and prevent this complication. This review suggests that robotics utilization improving TKA balancing has the potential to impact and reduce MFI. However, further research in this area is essential to provide insight regarding the role of robotics in mitigating the MFI risk.

1. Introduction

Total knee arthroplasty (TKA) is the most common surgical procedure for end-stage knee osteoarthritis [1,2,3]. It is increasing worldwide, with an annual TKA incidence boost of 119% from 2000 to 2014 per 100,000 population and an estimated growth of 84.9% to 1.26 million procedures by 2030 in the U.S. [2,3]. Despite evolving implants and technologies, about 20% of patients report dissatisfaction after TKA [2,4]. Experts agree that TKA success is based on precise implant placement, proper alignment, effective patellar tracking, and meticulous soft tissue balancing [2,5]. Robotic TKA (rTKA) techniques have emerged to improve TKA results and patients’ satisfaction, employing tools to execute precise bone cuts, improve components’ alignment, and offer optimal balancing [2,6,7,8]. The knee balance throughout its entire range of motion (ROM) plays a key role, and joint stability emerges as an essential determinant of TKA success [4,6]. However, despite technological advancements, instability remains a persistent challenge [6]. Mid-flexion instability (MFI), a specific instability form, has been described as a distinct clinical entity in TKA [1,9,10] and has been reported as a patient dissatisfaction cause [1,4,9]. Primarily introduced by Martin and Whiteside on cadaveric knees, it consists of instability in the mid-range of knee motion, with knee stability in full extension and flexion [9,10,11].

While TKA flexion and extension instability have been well investigated and associated with several treatment algorithms, the MFI concept remains poorly understood and ambiguous. This is because there are no universal diagnostic criteria, and patients report vague symptoms such as knee instability, recurrent effusion, challenges in daily activities, and pain [1,9].

Yoon et al., in their study, reported that 36% of patients exhibited laxity in mid-flexion during intraoperative stability testing of TKA despite a rectangular flexion and extension gaps achieved in gap balancing [12]. Consequently, obtaining joint stability within the mid-flexion range becomes crucial for optimal TKA functional outcomes [4,12].

MFI causes are various [1,9]. Vajapey et al. identified technique-specific, implant-specific, and patient-specific risk factors [9]. While some authors have revealed an association with mid-flexion instability, others give inconclusive and contradictory results [1,9]. Since robotics modifies factors affecting MFI, the question remains how their use could impact the MFI concept, possibly avoiding it [6,13,14,15].

This study aims to elucidate the MFI concept through a comprehensive literature review. First, it describes its definition, diagnosis, and risk factors, briefly introducing robotics and its evolution over time. Then, it explicitly attempts to investigate how robotic surgery, compared to manual TKA surgical techniques, could impact the MFI concept.

2. Search Strategy

A comprehensive analysis of the current literature regarding MFI was conducted. The research was performed using the PubMed and Scopus databases, with no limitations about study designs or level of evidence. No time limitation was used, but the most updated literature, mainly from 2019 to 2024, was analyzed. Several aspects of MFI were summarized and described in separate paragraphs, focusing primarily on its relationship with robotic surgery TKA.

3. Definition, Mechanism, and Patient Dissatisfaction of MFI

Martin and Whiteside introduced the MFI concept in 1990 [11] when the authors described an increased laxity during mid-flexion in ten cadaveric knees. MFI represents a distinct clinical phenomenon, characterized by stability at full extension and 90° flexion but instability during mid-range flexion, occurring between 0° and less than 90° [9,10,11]. In detail, MFI was identified by laxity in the coronal plane (varus/valgus) between 30° and 60° of knee flexion; it should be underlined that losses in the sagittal (anterior/posterior) and axial (rotational) planes may also occur within this mid-flexion range [9,11]. As recently reported by Longo et al., the term mid-flexion, as used in TKA literature, includes a variety of definitions, highlighting the lack of consensus on its precise flexion range [1].

Shalhoub et al. reported that a conventional gap balancing with equal flexion and extension gaps would result in approximately 2–4 mm residual laxity in mid-flexion despite a balanced knee in flexion and extension [16]. This scenario may be related to soft tissues, particularly medial collateral ligament (MCL) anatomy; they are tightest at full extension. However, as the knee flexes, its insertion distance gradually approximates, reducing MCL load and increasing laxity [16] (Figure 1).

MFI is a cause of patient dissatisfaction, as reported in the literature [4,17]. Patient satisfaction is an essential outcome following TKA, and it is well known that there is a significant difference between patient satisfaction and expectations and surgeon clinical outcomes evaluation [4,17]. For this reason, the new 2011 Knee Society Knee Scoring System (KSS 2011) includes patients’ perspectives [4,17,18]. Mid-flexion medial laxity has been associated with worse expectations because, during daily living activities, the knee is utilized in mid-flexion and not in full extension [4,17]. Nonetheless, as reported by Vajapey et al., MFI assessed intraoperatively is not always symptomatic postoperatively [9].

4. MFI Diagnosis

No standardized criterion for MFI diagnosing is described, and a universal definition does not exist [1,9]. For this reason, as reported by Vajapey et al., the diagnosis could be challenging due to its vague symptoms and signs [9].

Individuals with MFI could report difficulties in performing daily activities [9]. Additionally, recurrent joint effusions, diffuse periarticular pain, and overall dissatisfaction should prompt consideration of MFI by the examiner [1,9,19]. Radiographs, advanced imaging, and laboratory tests do not have a defined role in MFI diagnosing; for this reason, careful collection of patient data and preoperative and postoperative radiographs may help identify risk factors that predispose to MFI [9,11]. Radiographic imaging after TKA is performed, including full-length standing anteroposterior, skyline, and lateral radiographs [9,19]. Any deviations in alignment, fixation, or prosthetic components/patella positioning should be carefully examined [1,9,19]. For example, the posterior femoral condylar offset should be scrutinized and restored since its alteration may heighten the MFI risk [1,9,20,21].

In addition, apart from in the case of rTKA, certain variables assessed preoperatively, such as an initial coronal deformity > 10° and a preoperative joint line convergence angle (JLCA) value exceeding 2°, could assist the surgeon in determining the necessity of sensors helping devices intraoperatively to obtain a balanced knee [22,23].

This analysis underscores the complexity of diagnosing and managing MFI in TKA [1,9]. It highlights the need for a thorough understanding of knee mechanics and for identifying the risk factors to prevent potential MFI [1,9].

5. Risk Factors for MFI

Risk factors (RFs) for MFI have been categorized as implant-related, technique-related, and patient-related [1,9]. They are summarized in

Table 1. Mid-flexion instability risk factors. TKA: total knee arthroplasty; CR: cruciate-retaining; PS: posterior-stabilized; PCO: posterior condylar offset; °: degree.

Patient-Specific Risk Factors [1,9]	Implant-Specific Risk Factors [1,9]	Technique-Specific Risk Factors [1,9]
Preoperative laxity Type of patient activity after TKA	Conformity of the articular surface CR vs. PS prosthesis design Radius-of-curvature of the femoral component Bicruciate-retaining prosthesis design Femoral component size Bearing type in PS knees	Joint line elevation Type of alignment (mechanical versus anatomic) Positioning of the femoral component Type of gap balancing method Restoration of PCO Gap balancing at flexion 90° and extension 0°

.

Vajapey et al., in a recent systematic review and meta-analysis of 18 studies, identified most RFs as implant and technique-related, respectively [9]. They reported five of them to have contradictory results in the literature with no obvious association with MFI, such as the joint line position and implant design utilizing posterior-stabilized (PS) versus cruciate-retaining (CR) techniques [9]. Moreover, they supported that the articular surface conformity and preoperative joint laxity may play a more significant role than earlier believed [9]. Patient-related factors, such as the patient’s laxity and activity level of the patient after TKA, were described as contributors to MFI [9].

Some studies analyzed RFs, such as joint line position, specifically [1,9,21,24]. They reported that joint line elevation increases MFI [24] (Figure 2).

At the same time, Matziolis et al. found that proximalization by 5 mm or distalization by 2 mm of the joint line had no significant effect on MFI [21]. The systematic review by Longo et al. suggested that joint line position can be altered by up to 5 mm without measurable differences in joint stability [1]. Contradictive results are reported concerning the implant design utilized [1,9,24,25,26,27]. Hino et al. revealed that PS TKA increased joint laxity between 10° and 20° of flexion and that PS TKA provided more joint laxity than CR TKA throughout the whole ROM [25]. This conclusion could be explained by anatomical studies by Mihalko et al., where the primary result of posterior cruciate ligament sacrifice was creating a larger flexion gap [28]. However, Evangelista et al. demonstrated no difference in the mid-flexion range between PS and CR implants [24]. Interestingly, examining CR TKA, Tsuda et al. compared mid-flexion rotational laxity between two models of CR TKA: symmetrical surface design with neutral joint line obliquity and asymmetrical surface design with varus joint line obliquity [29]. They demonstrated that asymmetrical surface design reduced rotational laxity at the mid-flexion range in CR TKA [29].

Recently, attention has emerged towards the conformity of the articular surface, with high-conformity knees leading to decreased anterior femoral translation and reduced MFI [9,30]. In this regard, the same authors underlined that paradoxical anterior femoral translation (AFT) relies not only on the articular conformity of the prosthesis but also on the patient activity, as previously cited, since in low-conformity knees, the sit-to-stand activity leads to the highest level of paradoxical AFT, in moderate-conformity knees the crossover and pivot turn held the same outcome being the worst-case activity, and finally, all activities produced a relatively small AFT in high conformity knees [30].

Additionally, posterior condylar offset (PCO) influences MFI [1,9]. Matziolis et al. described how upsizing and downsizing the femoral component changes the PCO, leading to MFI [21].

Concerning the radius-of-curvature of the femoral component and MFI, results are contradictory [1,9,31]. Despite introducing a single-radius design to avoid transient ligament laxity and instability during knee mid-flexion, Stoddard et al. did not demonstrate enhanced mid-range stability of this type of implant over the multi-radius of curvature design. Consequently, they suggested the unknown ligament laxity during surgery as driving to mid-range instability [31].

As reported by Vajapey et al., the impact of TKA alignment on MFI is controversial [9,32,33]. Incavo et al. supported anatomic alignment leading to lateral joint space laxity in the mid-flexion range. In contrast, mechanical alignment was associated with a lateral joint space opening by 2–3 mm throughout the knee range of motion [32]. In contrast, in another study, no difference in coronal plane stability in the mid-flexion range was seen between kinematically or mechanically aligned cadaveric knees [33].

6. Evolution of TKA Surgery and Principles of Robotic TKA

TKA implants and instrumentations have evolved [2,34]. Recent implant innovations include ultra-congruent prosthesis designs [35], and knee prosthetic surgery instrumentation evolved from elementary tools in the early 1970s to navigation systems and rTKA [2,34,36].

Saragaglia et al. deeply analyzed the computer navigation systems and reported that they had not achieved the success they merited [34]. Conventional TKA is based on manually positioned jigs according to radiographs and intraoperative anatomical landmarks; on the contrary, rTKA utilizes computer software to convert anatomical information into a virtual patient-specific 3D reconstruction of the knee joint [2,36,37].

This advancement enables the customization of knee prosthesis alignment to replicate individual anatomy [2,36,37]. Robotic surgery exhibits significant potential in enhancing component placement precision, reducing complication rates, and ensuring better soft tissue preservation, offering real-time feedback regarding bone cut thickness or orientation [2,6,7,8,36,37].

Robotic-assisted surgeries provide several options for the surgeon, with robotic arms classified as fully active, passive, and semi-active [2,36]. As reported by Cantivalli et al., the three most commonly used robotic systems for knee arthroplasty are all semi-active [2]. They allow the surgeon to control bone resection and acquire live intraoperative feedback about any variation from the surgical plan [2]. On the contrary, active robotic systems function autonomously and complete the resections based on the surgeon’s planning, while passive systems are continuously controlled by surgeons [2]. Another important aspect involves “closed” or “open” platform systems, meaning the possibility to use only one specific implant or various implant designs [2].

Despite the initial economic cost and the increased operative time during the learning curve, studies demonstrate that the precision and high accuracy obtained with rTKA may improve patient-reported outcomes (PROMs) and long-term survivorship [2,6,7,8,13,36,37]. Kayani et al. analyzed that rTKA reduced postoperative pain and improved early functional rehabilitation compared to conventional jigs-based TKA [36]. Mulpur et al. reported that robotic surgery improves patient satisfaction, with most patients suggesting rTKA to others [14].

However, the two techniques had no differences in medium-to-long-term functional outcomes [2,14,36].

The literature emphasized that, while recognizing the potential benefits of rTKA, it has yet to acquire endorsement among orthopedic surgeons [2]. Obstacles are associated with conflicts of interest, problems in intraoperative registration, increased radiation exposure, a long intraoperative time during the learning curve, and the cost of robotics [2]. However, as reported by Cantivalli et al., robotics has been introduced in sports medicine for anterior cruciate ligament reconstruction and osteochondral lesions, representing a turning point for future knee surgery [2].

7. Treatment Strategies and Impact of Robotics on MFI

MFI is a critical concern in TKA, mainly when joint line elevation causes the flexion axis to no longer coincide with the MCL insertion. This instability may result from various surgical interventions and has significant implications for patient outcomes. Here, we discuss several strategies to mitigate MFI and the impact of robotic surgery on its management.

One surgical strategy to reduce the risk of MFI involves maintaining joint line height [1,9]. According to multiple authors, preserving the anatomical position of the joint line is crucial. Martin and Whiteside, who introduced the MFI concept, demonstrated that elevating the joint line by 5 mm in a proximal and anterior direction alters the femur’s rotation center relative to the collateral ligaments, leading to decreased tension at mid-flexion while maintaining stability in extension and at 90° of flexion [11]. Despite some debate [21], the joint line remains a fundamental aspect of knee stability [1,7,15].

Joint line elevation may result from additional distal femoral resection, a strategy sometimes used to address flexion contracture [38]. However, as Hardy et al. reported, each millimeter of additional femoral resection yields only a 2° improvement in extension, with less than 5° gained from a 2 mm resection [38]. Therefore, some authors suggest alternative techniques, such as posterior osteophyte resection or posterior capsular release, instead of relying solely on distal femoral resection [38].

rTKA has shown promise in reducing joint line deviation. Liow et al. found that the robot-assisted group had only 3.23% joint line outliers (>5 mm deviation) compared to 20.6% in the traditional group. However, no significant differences in short-term clinical outcomes were noted [7]. Agrawal et al. also demonstrated that rTKA achieved nearly anatomical joint line positions compared to conventional methods [15].

Another strategy to address MFI is using prostheses with a single-radius design rather than a multi-radius design [1,9,39]. Wang et al.’s kinematic study revealed mid-lateral valgus/varus instability at mid-flexion in knees with a multi-radius design during sit-to-stand movements. In contrast, a single-radius design provided better stability [39]. Gordon et al. suggested combining robotic surgery with intraoperative sensor feedback enhances knee balance with a single-radius implant design, achieving consistent stability throughout the mid-flexion range [40].

Despite ongoing debates about the influence of TKA alignment on MFI, recent systematic reviews and meta-analyses by Alrajeb et al. indicated that robotic knees had significantly better postoperative anatomical and mechanical alignment than conventional methods [13].

Classical versus reversed gap balancing techniques are also discussed [1,9]. Matziolis et al. introduced the reversed gap balancing technique, which aligns the femur according to the surgical transepicondylar axis and adjusts the femoral and tibial components based on the extension and flexion gap [41]. This technique reduces instability at various flexion angles, making soft tissue releases less extensive [41].

Regarding the impact of robotic systems on knee balance and loads post-TKA, Held et al. found that rTKA improved intraoperative compartment balancing in flexion, with no significant differences in mid-flexion and extension compared to traditional TKA [6]. Gordon et al. further demonstrated that combining robotics with real-time intraoperative sensor feedback increased the proportion of balanced knees from 67% to 87% by allowing for a wide range of soft tissue and bone adjustments [40].

Lastly, differences between robotic systems were evaluated [42,43,44]. For instance, Hasegawa et al. compared early outcomes and cutting accuracy using two different systems and found variations in medial and lateral laxities at 30°, likely influenced by the specific prosthesis designs utilized by each system [42].

8. Clinical Results and Evaluation

It is well established that prosthetic components’ malalignment and soft tissue imbalance primarily contribute to patient dissatisfaction with mediolateral laxity at 60° and MFI, leading to negative patient expectations [4,5] (Figure 3).

Robotics were also developed to solve this problem [2,36]. Even if their increased precision and knee balancing are well established [2,6,7,8,36], it remains to be proven if their high accuracy leads to better PROMs [2,14,36].

As already mentioned, Mulpur et al. performed a study in which the same patient underwent a staged bilateral TKA with one conventional manual TKA and an rTKA to assess PROMs and patient satisfaction [14].

This study underlined higher patient satisfaction, a reduced time to independent ambulation, and improved joint perception during daily activities measured through the Forgotten Joint Score-12 (FJS-12) with rTKA [14]. Patients felt the robotic-operated knee was less painful and more natural, and when asked which procedure they would prefer for surgery again, the majority preferred robotic [14]. Furthermore, 76.4% of the patients would recommend rTKA to others who should undergo knee surgery for osteoarthritis [14].

However, manual and robotic TKA were associated with comparable PROMs evaluated through the Oxford Knee Score [14]. Kayani et al., in a recent systematic review, described no differences in medium- and long-term functional outcomes between manual TKA and rTKA [36].

Wakelin et al. conducted a prospective study analyzing intraoperative balance and laxity in rTKA and their association with 2-year KOOS subscores [45]. The authors reported that intraoperatively measured balance and joint laxity are linked to PROMs at least two years after TKA [45]. Using these measurements, clinically laxity and balance limits were recognized throughout flexion, isolating a subgroup of TKAs with significantly improved pain outcomes [45].

9. Challenges, Limitations, and Future Perspectives

With new insight into MFI, some concerns remain [1,9]. First, few studies, such as clinical trials, investigate MFI after both conventional and rTKA [1,9]; moreover, most are laboratory-based or cadaveric [1,11]. As reported by several authors, this heterogeneity is probably due to the absence of accurate diagnostic criteria because MFI diagnosis is insidious due to the presentation of subtle symptoms [1,9]. While more recent papers agree to identify MFI as a distinct entity, some studies questioned whether it is separate from flexion and extension instability [1,9]. In addition, MFI is usually assessed intraoperatively without a preoperative evaluation, and it is essential to underline that this entity does not always lead to postoperative symptoms [9]. Due to the lack of diagnosis, treatment algorithms are not present, so surgeons are trying to mainly address the RFs [1,9,15].

Novel technologies such as robotics have gained prominence in contemporary surgical settings to enhance soft tissue balance and prosthetic component placement precision [2,6,7,8,13,14,15,46]. Through rTKA, robotics are addressing and improving RFs, potentially reducing the risk of MFI [2,6,7,8,13,14,15]. This promising development, however, is not without its challenges. Even in this field, direct studies analyzing mid-flexion are lacking [2,6,7,8,13,14,15]. In this regard, Agrawal et al. reported an almost anatomical restoration of the joint line position through robotics concerning the conventional method, but only indirectly citing the MFI concept [15]. The adoption of robotics in TKA remains a topic of debate, with an inconclusive demonstration of clear benefits in PROMs [2,6,7,8,13,36]. Moreover, incorporating robotics in the operating room means higher costs than traditional techniques [2]. Despite these challenges, the potential of robotics in addressing MFI and improving patient outcomes is a reason for optimism in the field of TKA surgeries.

Mulpur et al. analyzed patients experiencing staged bilateral knee arthroplasty, with one TKA performed with a manual approach and the other with a robotic one [14]. Despite comparable PROMs, rTKA was perceived as more natural and made patients satisfied or very satisfied [14]. For this reason, patients would recommend rTKA to other individuals [14].

A significant issue in the current literature [32,34,37,41] on rTKA is the lack of distinction between robotic systems, particularly image-based versus image-less systems, and various alignment strategies [37,41,46]. These factors could play a role in the occurrence of MFI. Image-based systems utilize preoperative imaging to guide the surgery, potentially offering more precise alignment and component positioning. In contrast, image-less systems rely on intraoperative data and may vary in accuracy [32,41,46]. Similarly, strategies such as mechanical, kinematic, and functional alignment have different implications for knee joint stability and the risk of MFI [32,34,42,43]. Future studies comparing these variables could provide essential insights into better understanding the impact of rTKA on MFI and improving surgical outcomes.

Finally, studies need to clarify diagnostic criteria to make the literature more uniform regarding MFI. Indeed, starting from a definite diagnosis, a correct management algorithm could be performed. Risk factors (RFs) should be evaluated, and their possible impact on MFI should be carefully analyzed. Moreover, MFI should be investigated as a parameter during the surgical procedure and as a postoperative outcome [9]. New risk factors could also be introduced; for instance, Vajapey et al. presented a potential unexplored factor requiring investigation: the patellar resurfacing effect on MFI [9]. Identifying patients at risk of MFI before surgery is crucial to prevent this outcome by utilizing the most appropriate techniques and implants. Further research in this area with higher-quality studies is essential to provide more clarity on MFI [9].

10. Conclusions

Despite technological advancements with robotic-assisted surgery, instability remains challenging in TKA. MFI, a specific form of instability in the mid-range of knee motion, has gained recognition as a distinct clinical entity in TKA and has been reported as a cause of patient dissatisfaction. With no universal diagnostic criteria, the MFI concept is still ambiguous, and no specific treatment algorithm is defined. For this reason, it is crucial to prevent and treat it by recognizing RFs, which could be technique-related, implant-related, or patient-related. Since robotics offers optimal balancing in TKA and reduces causes affecting MFI, robotics could indirectly impact and prevent this complication.