A proposal new classication in multi-ligament knee injuries based on both femorotibial joint and patellofemoral joint factors

Background Knee extensor structure disruption is sometimes combined with multi-ligament knee injury (MLKI). These cases could not be classied into the previous classication for knee dislocation or MLKI. To propose a new classication of MLKI that addresses both femorotibial and patellofemoral (PF) factors. The present study involved patients diagnosed with MLKI at our institute. Cases with two or more injured structures diagnosed clinically and by MRI and X-ray examinations were included. Details of injured structures were dened as follows: ‘cruciate ligaments’, including the ACL and PCL; ‘collateral ligaments’, including the MCL and the posterolateral corner (PLC); and the ‘PF joint factor’, including the medial patellofemoral ligament (MPFL), quadriceps tendon, and patellar tendon. A total of 65 cases met the inclusion criteria from 2007 to August 2020 and were enrolled in this study. and PCL+ patellar tendon (Type-A4), ACL+PLC with neurovascular injury (Type-A1-Y), and three cases of PCL+PLC with complications (Type-A1-X, -Y, -Z). With the new classication for MLKI and complications, all cases were successfully assigned to each category. In the present classication, every combination of MLKI cases containing both tibiofemoral and patellofemoral factors, with or without complications, was successfully classied into a suitable category. around the knee joint. Cases of two or more injured structures diagnosed clinically by the presence of the anterior/posterior drawer sign, Lachman test and pivot shift test, abnormal varus/valgus laxity, dial testing, and patellar subluxation, instability, and dislocation, as well as by magnetic resonance imaging (MRI), computed tomography (CT), and dynamic X-ray testing, were included. Cases with any prior knee surgery were excluded. A total of sixty-ve MLKI cases visited our institute and were provided surgical treatment due to knee instability and disability during the investigation period. No case had previous knee surgery in this series; therefore, all sixty-ve patients met the inclusion criteria and were enrolled in this study. cases are rare in part because the mechanism of injury was dicult to explain. classication of PF rare combinations remained in the categories.


Introduction
Multi-ligament knee injuries (MLKIs) occur as a result of high-energy trauma such as tra c accidents, fall accidents, and sports accidents. MLKIs show varied symptoms that depend on the combination of injured structures moreover, with additional periarticular fractures, nerve injury, and vascular injury in some MLKI cases [1]. MLKIs induce severe knee joint instability and leg disability. Therefore, careful assessment or examination is required for the diagnosis and treatment of MLKI [2].
The knee extensor mechanism, which consists of the quadriceps tendon, patella, and patellar tendon, is one of the important factors that determine knee function. Dislocation of the patella or quadriceps or patellar tendon rupture causing severe instability or disability of knee function. Of course, such injuries are also the part of the treatment regimen when considering treatment, be it repair or reconstruction. Sometimes, the knee extensor structure disruption is combined with injuries of other knee ligaments [3,4].
Several classi cation systems have been reported for the diagnosis of knee dislocation and MLKI cases. Kennedy's classi cation was focused on dislocation direction of the knee [5]. The French Society of Orthopedic Surgery and Traumatology (SOFCOT) 2008 classi cation classi ed into four categories based on the two main concepts of the mechanism of knee injury: gapping and translation [6]. These classi cations were focused on the femorotibial joint. Schenck's knee dislocation (KD) classi cation was also one of the great classi cations of knee dislocation and MLKI for surgeons with respect to decision-making [7,8]. Basically, KD classi cation was based on the combination of injured ligaments with or without the presence of fracture. However, KD classi cation also referred only to the cruciate and collateral ligaments. Therefore, it was reported that the KD classi cation in isolation was not predictive of any surgical management strategy [9]. Cases of combined cruciate or collateral ligament injuries with knee extensor structure disruption could not be classi ed in these previous classi cations. For example, the anterior cruciate ligament (ACL) and patellar tendon or medial patellofemoral ligament (MPFL) rupture cases that showed knee joint instability and disability and surgeons must consider treatment for both involved structures, were not included in the previous classi cations, even though two structures around the knee joint had failed.
The information about the number and combination of injured structures may affect surgical procedures and strategy with regard to the graft selection, surgical timing, surgical procedure and so on. Therefore, we proposed a new classi cation for MLKI to cover all combination of collateral ligaments, cruciate ligaments, and knee extensor structure in other words, patellofemoral (PF) joint factor.
Based on KD classi cation, we established and de ned several categories in accordance with the number and combination of injured structures, and additional combined injuries such as periarticular fractures, nerve injury, and vascular injury. It was hypothesized that all cases with at least two structures involved, not only a combination of femorotibial joint factors, but also PF joint factors and complications, could be classi ed using the new classi cation.

Patients
The present study was conducted in 2019, involving patients who were diagnosed with MLKI at our institute. The experimental design was reviewed and approved (Approval number No. O-0602) by the Ethics Committee of our institute. The procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2013. Information regarding the conduct of this study was disclosed, and research subjects (cases 1-61) were given the opportunity to refuse inclusion in this study. Written, informed consent was obtained from all patients and guardians of prospective research subjects (cases 62-65) for publication of this report and any accompanying images before entering their entry into this study. The patients who did not want to take part were not enrolled in this study. The study involved both a retrospective and a prospective observational design, with data collected from May 2007 to August 2020.
Basically, three senior orthopedic surgeons performed full physical examinations of each structure around the knee joint. Cases of two or more injured structures diagnosed clinically by the presence of the anterior/posterior drawer sign, Lachman test and pivot shift test, abnormal varus/valgus laxity, dial testing, and patellar subluxation, instability, and dislocation, as well as by magnetic resonance imaging (MRI), computed tomography (CT), and dynamic X-ray testing, were included. Cases with any prior knee surgery were excluded. A total of sixty-ve MLKI cases visited our institute and were provided surgical treatment due to knee instability and disability during the investigation period. No case had previous knee surgery in this series; therefore, all sixty-ve patients met the inclusion criteria and were enrolled in this study.

Establishment of new classi cation criteria
This study focused on the number and the combination of injured structures; both femorotibial joint and PF joint factors, and complications such as fractures and neurovascular injury. Based on the number of injured structures, cases with two injured structures were categorized as Type A, with three structures were categorized as Type B, with four structures were categorized as Type C, and with ve structures case were categorized as Type D. Injured structure counting was de ned as follows: cruciate ligament group, ACL and/or posterior cruciate ligament (PCL); collateral ligament group, medial collateral ligament (MCL) and/or posterolateral corner (PLC); and PF joint group, one of the MPFL or patellar tendon or quadriceps tendon. PLC contained several structures, certainly include the lateral collateral ligament (LCL), popliteus and popliteo-bular ligament and so on. The primary restrain structure for varus force is LCL, and for posterolateral rotatory instability is femoropopliteal complex. Sometimes not only LCL, but also popliteo-bular ligament is injured together. In the present study, LCL and popliteo-bular ligament were de ned as one category. The PF joint group mainly contained three different structures. However, these structures work together as the knee extensor mechanism with a central focus on the patella. Therefore, these PF joint factors were de ned as the same group, and it was rare that two or more structures failed at the same time in this group. Thus, when one of these structures was ruptured, this was counted as PF joint factor failure in this classi cation.
Depending on the combination of injured structures, each case was subdivided into 1 to 5 in Types A and B and 1 to 3 in Type C. For example, a case with one cruciate ligament and one collateral ligament injury was de ned as Type A-1, bi-cruciate ligament injury was de ned as Type A-2, both collateral ligament injuries were de ned as Type A-3, one cruciate ligament and PF joint factor failure was de ned as Type A-4, and one collateral ligament and PF joint factor failure was de ned as Type A-5. Type D contained ve injured structures (bi-cruciate ligaments, both collateral ligaments, and the PF joint factor); therefore, Type D was not subdivided. Additional complications were also subdivided as follows: MLKI with periarticular fracture was de ned as X, with neurovascular injury as Y, and both periarticular fracture and neurovascular injury as Z. For example, one cruciate ligament injury and one collateral ligament injury and an additional fracture were classi ed as Type A-1-X. Details of each category of the new classi cation are provided in Tables 1, 2, 3, and 4, respectively.   For each case, the nal decision regarding the injured structures around the knee joint and complications such as periarticular fracture and neurovascular injury were recorded after careful clinical examination and imaging evaluation including MRI, dynamic X-ray testing, ultrasonography, and CT scan with angiography for detecting vascular injury or minor periarticular fractures, if needed. Periarticular nerve ndings were also carefully evaluated. Not only complete paralysis of the lower leg including the foot and toe, but also slight numbness, sensory disturbance, and muscle weakness were evaluated to determine nerve injury. MLKI cases were divided into both KD classi cation and the present classi cation.

Results
Of the total of sixty-ve cases, the causes of MLKI were as follows: sixteen tra c accidents, seven fall accidents, four work-related accidents, four involved slipping and twisting, and thirty-four sports accidents. The sports were: volleyball in six cases; judo and soccer in ve cases each; baseball, basketball, and rugby football in three cases each; gymnastics and martial arts in two cases each; and wrestling, skiing, badminton, sur ng, American football in one case each. Of these sixty-ve cases, fty-one were divided by KD ACL+PLC with periarticular fracture as Type-A1-X; ACL+PLC with neurovascular injury as Type A1-Y; PCL+PLC with periarticular fracture as Type-A1-X; PCL+PLC with neurovascular injury as Type-A1-Y; PCL+PLC with periarticular fracture and neurovascular injury as Type-A1-Z; ACL+MCL+PLC as Type-B2; ACL+PCL+MCL+PLC+MPFL as Type-D; ACL+MPFL as Type-A4; PCL+PLC+ patellar tendon rupture with periarticular fracture as Type-B3-X; ACL+MCL+MPFL as Type-B3; and PCL+ patellar tendon as Type-A4 (Table 5).

Discussion
The most important nding of the present study was that all sixty-ve MLKI cases were successfully categorized by the new classi cation without confusion. On the other hand, fourteen cases (21.5%) could not be categorized using KD classi cation. Seven of these fourteen cases contained PF joint factor injuries, with MPFL rupture in ve cases and patellar tendon rupture in two cases. MPFL rupture due to patellar dislocation and patellar tendon rupture induce dysfunction and disability of the knee extensor system.
Disruption of the knee extensor mechanism are among the targets for treatment, such as reconstruction or repair [10]. For knee surgeons, how many structures are injured, and which structures are injured are key issues for determining the strategy for the graft selection, the surgical procedure, xation devices, xation order. Marx et al reported that preoperative planning is critical, and the order of tunnels, graft passage, and xation should be decided prior to surgery [11].
Knee dislocation or MLKI cases can become very challenging, and even more, sometimes limb-threatening due to neurovascular complications [12][13][14]. There have been several management protocols and algorithms for treatment of knee dislocation or MLKI cases reported previously [1,2,12,15]. Maslaris et al reported a comprehensive initial management protocol for knee dislocation or MLKI that was focused on emergency care taking into account pitfalls and potential complications [16]. Sheth et al suggested that early operative treatment of injured ligaments improved functional and clinical outcomes compared with conservative management or delayed surgery [17]. Fanelli et al concluded that surgical timing for MLKI depends on the ligaments injured, the vascular status of the extremity, reduction stability, and the overall health of the patient [12]. These ndings suggest that correct preoperative diagnosis of injured structures and complications, and recognition of the target for treatment are important for surgical planning. However, these previous reports addressed mainly femorotibial joint factors. The combination ruptures of PF joint factors and other ligaments have been reported. De Baere et al presented a case of a combined lesion of the MCL and patellar tendon rupture [18]. In this case, the patellar tendon was reconstructed with a fascia lata allograft, and the MCL was also reconstructed with a semitendinosus tendon autograft. Kim et al reported a case of combined injury of the ACL and patellar tendon; the patellar tendon was repaired with primary suture, and the ACL was reconstructed with an Achilles tendon allograft [19]. Levakos et al also reported six cases of combined injuries including ACL and patellar tendon ruptures [20]. In four of six cases, MCL rupture was also present. There is controversy regarding surgical treatment versus nonoperative options for the MPFL after a rst-time patellar dislocation [21]. Allen at al reported MPFL injury in MKLI cases. They concluded that an MPFL tear in MLKI rarely caused instability and does not need to be addressed at the time of injury [22]. On the other hand, Gormeli et al presented an MLKI case with ACL, MCL, and MPFL ruptures, and they performed reconstructive surgery for all three injured ligaments with a good clinical outcome [23]. In the present study, six of seven cases with combined PF joint factor injuries underwent surgical treatment due to instability and dysfunction (four for the MPFL and two for the patellar tendon). These ndings suggest that PF joint factor injury in MLKI cases may be considered a target of surgical treatment.
Additionally, KD classi cation, Kennedy's classi cation, and SOFCOT 2008 classi cation are based on the presence of dislocation. MLKI without a documented knee dislocation has often been described as a knee dislocation that 'spontaneously reduced` [24]. It is di cult and confusing to apply these previous classi cations for the already reduced cases before consultation with the physician in the hospital. MLKI without obvious dislocation or spontaneously reduced cases also provide severe instability and disfunction of the injured knee joint. Therefore, a suitable and systematic classi cation of MLKI regardless with or without dislocation is needed.
The cause of MLKI injury was categorized for high-energy injury and low-energy injury [25]. Actually, the cause is multifactorial, as follows: direct or indirect force; and magnitude, speed, direction of the external force; rotational or exion angle of the joints of the hip, knee, and ankle at the time of injury; and so on. Therefore, an inconceivable combination of injured structures can be seen in MLKI cases, sometimes. For example, ACL+MCL+PLC cases are rare in part because the mechanism of injury was di cult to explain. However, a case with this combination of injured ligaments was con rmed in this case series (case 51), and it was also unclassi able by KD classi cation. However, we sought to establish a new classi cation to cover all combination of collateral ligaments, cruciate ligaments, and PF joint factor. Therefore, the rare combinations were remained in the categories.
Complications such as periarticular fracture combined with MLKI were classi ed as KD-V by Schenck's KD classi cation [7,8]. However, the de ned categories were sometimes ambiguous, such as KD-V1, which was de ned as 'Fracture dislocation without cruciate ligament'. and KD-V2 was de ned as 'with isolated bi-cruciate lesion'. KD-V3, -4, and -5 are also based on a bi-cruciate lesion. From this description, it was unclear how many ligaments were ruptured. Eight periarticular fracture cases were seen in the present series. Five of eight cases were unclassi able by KD classi cation.; ACL + PLC with periarticular fracture (case 8 and 63), PCL + PLC with fracture (case 45 and 58), PCL +PLC +PF with fracture (case 35). All MLKI with periarticular fracture cases were assigned in the present classi cation regardless of bi-cruciate ligament injury and of without cruciate ligament injury.
For knee surgeons, comprehensive classi cation based on the severity and number of targets for surgery has been required. Therefore, we established and proposed a new classi cation for MLKI that includes every combination all femorotibial and PF joint factors and complications. In the present classi cation, every combination of MLKI cases with or without complications was successfully classi ed into a suitable category.

Limitations
Several limitations must be taken into consideration with respect to the present study. First, the study had a small sample size (sixtyve cases). Second, the MPFL, patellar tendon, and quadriceps tendon were categorized into one group as the 'PF joint factor group' and counted as one. Based on an extensive literature search, there have been no reports of cases that involved both patellar tendon and quadriceps tendon ruptures in the same knee at the same time, as well as no cases of quadriceps tendon rupture and MPFL combined injury. Only one case that involved patellar tendon rupture with injury of the MPFL in a oating knee with proximal tibial and distal femoral fractures has been presented [26]. Rupture of two or more structures of the PF joint factors at the same time seems quite rare in MLKI. Therefore, these structures were classi ed and counted as one group, such that PF joint factors are considered one knee extensor system in the present classi cation.

Conclusion
Disruption of PF joint factors were sometimes combined with other knee ligament ruptures in MLKI cases. Injury of least two structures was con rmed in these cases, and therefore they may be considered as tting into the MLKI category. However, these combinations were di cult to categorize by any previous classi cations for knee dislocation or MLKI. In the present classi cation, every combination of MLKI cases containing both femorotibial and PF joint factors with or without complications was successfully classi ed into a suitable category. regarding the conduct of this study was disclosed, and research subjects (cases 1-61) were given the opportunity to refuse inclusion in this study. Written, informed consent was obtained from all patients and guardians of prospective research subjects (cases 62-65) for publication of this report and any accompanying images before entering their entry into this study. The patients who did not want to take part were not enrolled in this study.

Author Contribution
Takuya Tajima participated to create the conception, design of the study, performed investigation and acquisition of data.
Yudai Morita, Makoto Nagasawa, Tomomi Ota and Shuichi Kawagoe made substantial contribution to acquisition of data, analyze, interpretation of data.
Takuji Yokoe and Nami Yamaguchi had been involved in drafting the manuscript, and revising it critically for important intellectual content.
Etsuo Chosa participated in the study design and coordination, given nal approval of the version to be published. All authors read and approved the nal manuscript.