Satisfactory Outcomes and Improved Range of Motion With Arthroscopic Lysis of Adhesions and Manipulation for Arthrofibrosis After Multiligamentous Knee Reconstruction

Abstract

Purpose

To (1) evaluate the efficacy of surgery for arthrofibrosis (AF), as measured by preoperative and postoperative range of motion (ROM), and (2) evaluate patient-reported outcomes at mid- to long-term follow-up.

Methods

We performed a retrospective review of a prospectively collected database including patients who sustained multiligamentous knee injuries (MLKIs) managed surgically, sustained loss of ROM after surgical intervention, and underwent subsequent lysis of adhesions (LOA) and/or manipulation under anesthesia (MUA). Loss of ROM was defined as clinically symptomatic loss of terminal extension (flexion deformity) and/or flexion compared with the contralateral side.

Results

In total, 12 patients (6 male and 6 female patients; age, 36.0 ± 8.7 years; body mass index, 36.3 ± 8.7) met the inclusion criteria and underwent LOA and/or MUA at a mean of 14 ± 27 months (median, 4.0 months; interquartile range, 3.5-9.3 months) after MLKI surgery. Prior to AF intervention, patients showed mean flexion of 75.9° ± 36.0° (range, 30°-129°), mean extension of 3.2° ± 5.2° (range, 0°-12°), and a mean arc of motion of 72.7° ± 34.1° (range, 30°-117°). At a mean follow-up of 7.0 ± 3.9 years (range, 2.4-16.6 years) after AF intervention, patients showed a significant increase in knee flexion of 49° (P = .003), a significant increase in arc of motion of 51° (P = .002), and an increase in extension of 3° (P = .086). The mean final International Knee Documentation Committee score was 59.5 ± 23.9; Lysholm score, 72.1 ± 20.6; Tegner activity scale score, 5.6 ± 2.8; visual analog scale score at rest, 1.0 ± 1.6; and visual analog scale score with use, 3.3 ± 2.5. At final follow-up, 2 patients (17%) had undergone conversion to total knee arthroplasty (TKA) at 10.3 and 24.8 years after MLKI surgery. Of the 10 patients who did not go on to TKA, 9 (90%) reported that they were satisfied or very satisfied with their AF knee surgery.

Conclusions

At mid-term follow-up, LOA and/or MUA for symptomatic AF after multiligamentous knee surgery results in high rates of patient satisfaction and improved knee ROM and pain scores, as well as durable and satisfactory functional outcomes in patients not undergoing TKA.

Level of Evidence

Level IV, therapeutic case series.

Arthrofibrosis (AF) after multiligamentous knee injury (MLKI) is well recognized but remains clinically challenging to treat.^1,2 Additionally, there continues to be controversy regarding the timing of surgery, reconstruction techniques and grafts, and use of 1- or 2-stage procedures.3, 4, 5, 6, 7, 8 Although this condition is poorly understood, studies have identified high rates of AF after MLKI ranging from 3% to 57%.^6,9,10 Proposed risk factors for the development of AF after MLKI have included decreased time from injury to surgery, greater degree of initial injury, and surgical complications such as graft malposition, infection, or synovitis.^6,11,12

The rationale for surgical lysis of adhesions (LOA) and/or manipulation under anesthesia (MUA) is to improve knee range of motion (ROM) and outcomes in MLKI patients. Treatment of AF after anterior cruciate ligament (ACL) reconstruction and total knee arthroplasty (TKA) has been shown to be efficacious.^13,14 In these populations, it is suggested that LOA removes extracellular matrix, which binds profibrotic factors such as transforming growth factor β,^11,15,16 as well as interrupts myofibroblast feedback loops and leads to apoptosis.¹⁷

However, to date, there remains a paucity of literature regarding the outcomes of surgical intervention for the treatment of AF after MLKI. A recent systematic review by Fahlbusch et al.¹⁸ identified only 86 patients across 25 studies from 1995-2021 who were treated for AF after MLKI, with a mean follow-up period of 47 ± 32 months. The purpose of this study was to (1) evaluate the efficacy of surgery for AF, as measured by preoperative and postoperative ROM, and (2) evaluate patient-reported outcomes (PROs) at mid- to long-term follow-up. We hypothesized that a significant increase in knee ROM with acceptable PROs would be observed.

Methods

After receiving institutional review board approval (No. 07-004018), we performed a retrospective review of a prospectively collected database of MLKIs. Patients who met the following criteria were included: knee dislocation (KD) or MLKI consisting of at least 2 of the 4 major knee ligaments (ACL, posterior cruciate ligament [PCL], medial collateral ligament/posteromedial corner, and lateral collateral ligament/posterolateral corner [PLC]), loss of ROM after MLKI, and subsequent intervention including LOA and/or MUA. Loss of ROM was defined as loss of terminal extension (flexion deformity) and/or flexion compared with the contralateral side. MLKI surgery, ROM interventions, and measurements were performed by 1 of 2 sports medicine fellowship–trained surgeons (B.A.L. and M.J.S.) at a single tertiary referral center. Patients were excluded if they had less than 2 years of follow-up.

The Schenck classification was used to determine the KD grade.^19,20 Surgical indications included the occurrence of any of the following after several months of physical therapy (PT), continuous passive motion (CPM) use, and/or dynamic (spring-loaded) bracing: (1) inability to achieve full extension or 5° shy of full extension, resulting in an abnormal gate; (2) inability to achieve greater than 90° of flexion; and (3) progressive loss of ROM after initial ROM gains. Surgical techniques for multiligamentous knee surgery varied over the study period at the discretion of the operating surgeon regarding repair or reconstruction—or a combination thereof. Graft selection as well varied and was based on surgeon preference. The multiligamentous surgical techniques performed have been previously published.^7,21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33

Surgical Technique for AF Surgery

Standard knee arthroscopy portals were used to complete diagnostic arthroscopy and identify any associated additional pathology. AF in cases of cyclops lesions or flexion contracture was addressed with anteromedial and anterolateral portals involving circumferential releases of both the medial and lateral retinacula, the suprapatellar pouch, and the anteromedial and anterolateral gutters. AF impeding full knee extension (flexion deformity) was addressed by removing any cyclops lesions or scar tissue from the intercondylar notch, and if full extension was not achieved, then both medial and lateral posterior capsular releases were performed through posteromedial and posterolateral portals.³⁴ Arthroscopic LOA was conducted using mechanical shavers, radiofrequency devices, and cutting instruments. For manipulation, the hip was flexed and the knee joint was loaded until a break of adhesions was felt and sufficient ROM returned. Usually, MUA was performed first, followed by LOA; however, in cases of severe AF, care was taken not to place excessive pressure through the knee joint and arthroscopic LOA was performed prior to MUA. Ten patients underwent arthrolysis and manipulation, whereas 2 underwent manipulation alone.

Multiligamentous Knee Reconstruction Rehabilitation Protocol

A standardized institutional multiligamentous knee reconstruction rehabilitation protocol was prescribed to all patients. In brief, after initial MLKI knee surgery, patients were placed in postoperative bracing locked in extension and limited to toe-touch weight bearing with crutches for 10 days. From day 10 to week 6, partial weight bearing (25%) with bracing locked in extension was allowed, with progression to weight bearing as tolerated and custom bracing 24 hours per day 7 days per week (this was continued throughout rehabilitation). From week 6 to week 36, progression of core, hip, and lower-extremity exercises was performed with avoidance of cutting, running, and jumping. Isolated hamstring exercises were also initiated at this time, unless a PCL or PLC injury was present, in which case such exercises were delayed until week 24. From week 36 to week 48, low-level sports activity was initiated with isokinetic testing.

AF Rehabilitation Protocol

After AF intervention, patients were reintroduced to their last point of progress along the previous multiligament knee rehabilitation protocol if they were still within the time frame. Although some variability existed, our standard protocol was to perform femoral nerve blocks with admission to the hospital for 48 hours for CPM and PT. Thereafter, the nerve catheters were removed, and patients were discharged home with CPM and PT prescribed 5 times per week for 3 weeks. Subsequently, depending on how patients were progressing, CPM was either continued for another 3 weeks or discontinued. The number of PT sessions per week was modified based on individual patient ROM progression. The medical management of AF at our institution has evolved over time. At present, management involves the following: intravenous dexamethasone prior to AF surgery incision; intraoperative corticosteroid injection; 1 dose of ketorolac immediately postoperatively, with subsequent administration every 6 hours, for a total of 4 doses; intravenous dexamethasone at 1 day postoperatively; oral aspirin twice a day for 6 weeks postoperatively; and celecoxib twice a day for up to 4 months postoperatively. MUA postoperatively may also be performed at 4 and 8 weeks; however, it was not performed in any of the patients in this cohort. Patients may not have undergone all of the aforementioned steps at the time of their intervention for AF.

Data Collection

Demographic data including age, sex, and body mass index were collected. Injury- and surgery-specific data including side, staging, high- or low-energy mechanism, timing of surgery, and presence or absence of meniscal, cartilage, peroneal nerve, and/or vascular injury, as well as KD grade, were also collected. The smoking status and diabetes status of patients were noted at the time of injury or after the injury. Postoperative ROM, International Knee Documentation Committee (IKDC) score, and Tegner activity scale score were also reported, the latter 2 of which have been validated for the evaluation of ligamentous knee injuries.35, 36, 37

Statistical Analysis

Statistical analysis was performed using BlueSky software (version 7.4.0; BlueSky Statistics, Chicago, IL). Continuous variables were expressed as mean ± standard deviation (range) or as median (interquartile range [IQR]) in the presence of outliers. The paired t test was used for pre- and post-lysis comparison. All statistical tests were 2-sided, with P < .05 considered statistically significant.

Results

Of the 345 patients in the prospectively collected MLKI database, 20 were identified to have met our inclusion criteria. Eight patients were lost to follow-up; thus, 12 patients (6 male and 6 female patients; age, 36.0 ± 8.7; body mass index, 36.3 ± 8.7) were included, with a mean follow-up period of 7.0 ± 3.9 years (range, 2.4-16.6 years) (Table 1). The patients underwent LOA and/or MUA at a mean of 14 ± 27 months (median, 4.0 months; IQR, 3.5-9.3 months; range, 36 days to 8.1 years) after MLKI surgery. The mean time between injury and MLKI surgery was 76 ± 149 days (median, 23 days; IQR, 11-71 days; range, 0-537 days). Regarding KD grade, 2 patients had grade IV injuries, 6 had grade III injuries (3 medial and 3 lateral), and 4 had grade I injuries. At baseline, 2 patients (17%) reported a history of smoking and 3 (25%) reported a history of diabetes. Prior to index surgery, all 12 patients underwent PT and were prescribed bracing 24 hours per day 7 days per week. Eight patients underwent allograft reconstruction alone, whereas 4 underwent both autograft and allograft reconstruction. Of note, peripheral nerve blocks were used in 7 patients (58%) and CPM machines were used by 6 patients (50%) after AF surgery. Prior to AF intervention, patients showed mean flexion of 75.9° ± 36.0° (range, 30°-129°), mean extension of 3.2° ± 5.2° (range, 0°-12°), and a mean total arc of motion of 72.7° ± 34.1° (range, 30°-117°). At a mean follow-up of 7.0 ± 3.9 years (range, 2.4-16.6 years) after AF intervention, patients showed mean flexion of 124.5° ± 9.1° (range, 110°-140°), mean extension of 0.5° ± 1.6° (range, 0°-5°), and a mean total arc of motion of 123.8° ± 8.9° (range, 110°-140°). Patients showed gains of 48.6° in flexion, 2.7° in extension, and 51.1° in the arc of motion. Full patient characteristics and mean ROM measurements are displayed in Tables 2 and 3, respectively.

Table 1.

Injury and Surgery Characteristics

Characteristic	Data
Mean age at index surgery, yr	36.0 ± 8.7
Side
Left	5 (41.6)
Right	7 (58.3)
Staged operation
Single-stage operation	10 (83.3)
2-Stage operation	2 (16.6)
Mechanism of injury
Low energy	6 (50.0)
High energy	6 (50.0)
Timing of initial MLKI surgery after injury
Acute (<21 d)	7 (58.3)
Delayed (>21 d)	5 (41.6)
Meniscal injury
Not present	7 (58.3)
Present	5 (41.6)
Cartilage injury
Not present	8 (66.6)
Present	4 (33.3)
Peroneal nerve injury
Not present	10 (83.3)
Present	2 (16.6)
Vascular (popliteal artery) injury
Not present	12 (100.0)
Present	0 (0.0)
KD grade per Schenck classification scheme19,20
I	4 (33.3)
II	0 (0.0)
III	6 (50.0)
IV	2 (16.6)
V	0 (0.0)

NOTE. Data are presented as mean ± standard deviation or number (percent of total).

KD, knee dislocation; MLKI, multiligamentous knee injury.

Table 2.

Patient Characteristics

	Ligaments Injured	Ligaments Repaired in Initial Surgery	Staged Operation	Time to LOA and/or MUA, mo	ROM, °		TKA
					Before Intervention	After Intervention
Patient 1	PCL, MCL	PCL, MCL		98	—	—	Yes
Patient 2	ACL, PCL, LCL	ACL, PCL, LCL		4	30	124
Patient 3	ACL, PCL, MCL	ACL, PCL, MCL		17	107	125
Patient 4	ACL, PCL, MCL, LCL	ACL, PCL, MCL		1	—	—	Yes
Patient 5	PCL, LCL	—	Yes	3	30	140
Patient 6	ACL, MCL	ACL, MCL		5	117	135
Patient 7	PCL, LCL	PCL, LCL		14	110	120
Patient 8	ACL, PCL, MCL	ACL, PCL, MCL		8	30	120
Patient 9	PCL, LCL	PCL, LCL		4	60	114
Patient 10	ACL, PCL, LCL	LCL	Yes	3	73	135
Patient 11	ACL, PCL, MCL	ACL, PCL, MCL		4	90	125
Patient 12	ACL, PCL, LCL	ACL, PCL, LCL		4	80	110
Mean ± SD or n (%)			2 (17)	14 ± 27	73 ± 34	124 ± 8.9	2 (17)

ACL, anterior cruciate ligament; LCL, lateral collateral ligament; LOA, lysis of adhesions; MCL, medial collateral ligament; MUA, manipulation under anesthesia; PCL, posterior cruciate ligament; SD, standard deviation; TKA, total knee arthroplasty.

Table 3.

Pair-wise Comparison of Preoperative and Postoperative Motion

Motion	Preoperative (n = 10)	Postoperative (n = 10)	Change	P Value
Flexion, °	75.9 ± 36.0 (30-129)	124.5 ± 9.1 (110-140)	48.6	.003∗
Extension, °	3.2 ± 5.2 (0-12)	0.5 ± 1.6 (0-5)	2.7	.086
Arc of motion, °	72.7 ± 34.1 (30-117)	123.8 ± 8.9 (110-140)	51.1	.002∗

NOTE. Data are presented as mean ± standard deviation (range).

^∗Statistically significant (P < .05).

PRO Scores

At a mean of 7 years’ follow-up after AF intervention, patients showed a mean IKDC score of 59.5 ± 23.9; Lysholm score, 72.1 ± 20.6; Tegner activity scale score, 5.6 ± 2.8; visual analog scale score at rest, 1.0 ± 1.6; and visual analog scale score with use, 3.3 ± 2.5.

Patient Satisfaction

Regarding satisfaction with surgery among patients who did not undergo TKA, 1 patient reported “neutral” satisfaction whereas 3 patients were “satisfied” and 6 patients were “very satisfied.” Overall, 90% of patients reported that they were satisfied or very satisfied with their AF knee surgery, and no patients reported being “dissatisfied” or “very dissatisfied.”

Reoperations and Complications

At final follow-up, 2 patients (17%) had undergone conversion to TKA at 10.3 and 24.8 years after MLKI surgery (10.2 and 16.6 years after AF intervention). In 2 patients who did not undergo TKA, subsequent, planned staged operations were performed at 5 months (ACL reconstruction with contralateral hamstring autograft) and 18 months (all-inside PCL reconstruction with tibialis allograft and PLC reconstruction) after MLKI surgery (2 and 15 months after AF intervention). The 2 patients who underwent staged operations had final arcs of motion of 0° to 135° and 0° to 140°, respectively, at final follow-up (Table 2). No complications were reported specifically due to arthrolysis or manipulation.

Discussion

The most important finding of this study was that patients undergoing MUA and arthroscopic LOA after MLKI have significant and durable improvements in flexion and the overall arc of motion compared with preoperative measurements. Satisfactory PROs and patient satisfaction were also observed. These results affirm our hypothesis.

The 51° increase in the overall arc of motion is similar to the findings of a systematic review by Fahlbusch et al.,¹⁸ who—when combining 5 studies consisting of 12 patients in total—described an increase in ROM of 48° ± 10° (range, 33°-60°) at 6.3 ± 3.0 months (range, 1.0-12.0 months) after LOA. Furthermore, a recent systematic review by Fackler et al.³⁸ showed a 42° improvement in the knee arc of motion after LOA and MUA across 8 studies comprising 240 patients with all-cause AF. Notably, the greater motion gained in our study may reflect more severe AF in MLKI patients, with a greater capacity for an increase in motion after LOA.^38,39

This study shows satisfactory final PROs as measured by IKDC, Tegner, Lysholm, and VAS scores, with 90% of those not undergoing TKA reporting that they were satisfied or very satisfied with their AF surgery. In patients treated for MLKI without AF, Engebretsen et al.⁴⁰ and Ranger et al.⁴¹ reported IKDC scores of 64 ± 20 and 68 ± 20, respectively, and Lysholm scores of 81 and 79 ± 19, respectively. Other studies have reported similar results.^42,43 Levy et al.⁴⁴ reported an IKDC score of 62 and Lysholm score of 69 at median 5-year follow-up in 64 patients older than 30 years. In their study, the 61 patients aged 30 years or younger had significantly higher IKDC and Lysholm scores of 73.3 and 76.9, respectively. Thus, the older average age of our cohort (i.e., 36 years) may contribute to the modest mean IKDC score of 59 and Lysholm score of 72.

Of note, our study results in patients treated with LOA and/or MUA for AF after MLKI are overall similar to those reported by prior authors for isolated MLKI (i.e., no associated intervention for AF), suggesting the safety and efficacy of AF intervention. The strengths of this study include the use of a prospectively generated MLKI patient database and the strict inclusion criteria including MLKI surgery and definitive arthroscopic AF surgery.

Limitations

The primary limitation of the study is the 40% rate of overall loss to follow-up, which introduces the possibility of nonresponse bias.⁴⁵ Furthermore, the small number of patients limits the power to perform subgroup analysis. Additionally, the cohort is heterogeneous regarding the KD grade, timing of surgery, surgical techniques performed, and use of a single- versus 2-stage approach. That said, this heterogeneity is reflective of most MLKI cohort studies. Finally, the strict inclusion criteria do not account for patients who had AF and were either managed nonoperatively or treated medically.

Conclusions

At mid-term follow-up, LOA and/or MUA for symptomatic AF after multiligamentous knee surgery results in high rates of patient satisfaction and improved knee ROM and pain scores, as well as durable and satisfactory functional outcomes in patients not undergoing TKA.