Polyethylene exchange versus full component revision for arthrofibrosis following total knee arthroplasty: A retrospective cohort analysis

Abstract

Introduction

Arthrofibrosis occurs in up to 10% of patients who undergo total knee arthroplasty (TKA). For cases that are not amenable to manipulation under anesthesia (MUA), there is little consensus on how many components should be revised. We compared outcomes of polyethylene exchanges, when doing lysis of adhesions (LOA) versus full component revision. Specifically, we assessed: (1) re-operation rates at one and two years; (2) rates of surgical complications at one year; and (3) associated risk factors for requiring a LOA.

Methods

A database queried all patients who underwent revision TKA for arthrofibrosis. A total of 2410 patients were identified, 1120 (46.5%) of which underwent all-component knee revision, while the remaining 1290 (53.5%) underwent polyethylene spacer revision. Multivariate logistic regressions assessed re-operation rates and risk factors for requiring LOA.

Results

The incidence and odds of re-operation within one year following polyethylene exchange was not significantly different than full component revision (10.9 versus 12.9%, odds ratio (OR) 0.83, 95% confidence interval (CI) [0.64–1.06], p = 0.145). However, the adjusted models for re-operation within one- and two-years following LOA in the form of polyethylene exchange revision was significantly higher than the full component revision cohort (OR 1.52 CI [1.07–2.17], p = 0.022 and OR 1.44 CI [1.06–1.97] p = 0.022). Risk factors associated with the need for lysis of adhesions included age less than 60 years, depression, fibromyalgia, and anxiety.

Conclusions

Full component revision TKA for arthrofibrosis was associated with lower two-year re-operation rate than polyethylene exchange. Risk factors for LOA include younger age and fibromyalgia.

1. Introduction

Arthrofibrosis, which is characterized by an overactive inflammatory response and proliferation of fibroblasts leading to excessive scar tissue, occurs at a rate of approximately 5% following a primary total knee arthroplasty (TKA).^1,2 It is a debilitating complication associated with poor patient outcomes and satisfaction.^3,4 It is also one of the leading causes for return to the operating room within the first 3 months following a primary TKA.1, 2, 3, 4, 5 The most effective nonsurgical methods of prevention includes combinations of patient education, aggressive postoperative physical therapy, and the utilization of anti-inflammatories.⁶ When these techniques are unsuccessful, then surgical management is utilized. These include manipulation under anesthesia (MUA), arthroscopic versus open lysis of adhesions, and revision arthroplasty.

MUA is oftentimes utilized in the first 6–12 weeks following a knee arthroplasty in patients who fail to progress in range of motion (ROM). Earlier use has been associated with improved patient satisfaction as well as postoperative outcomes.⁷ Crawford et al. reported that patients who undergo manipulation under anesthesia after primary total knee arthroplasty have an increased risk for revisions, worse long-term clinical scores, and poor survivorships.^8,9 An open lysis of adhesion (LOA) procedure with polyethylene exchange, allows for direct visualization and treatment of pathologic fibrous scar tissue when MUA has failed to restore range of motion.¹⁰ Previous studies have showed that polyethylene exchange with arthrolysis can improve flexion 10–34°, and extension 12–23°, with an average overall improvement of 15–40°. Despite this, partial component revision has been reported to have a 23.8% revision rate within two years.^10,11 Additionally, the utility of this technique has been questioned since gains in range of motion have been comparable to MUA and may exist only for patients who have severe arthrofibrosis.²

All component revision has gained in frequency and popularity in recent years. An advantage of this technique is that it addresses any and all issues as a pertains to potential malalignment and soft tissue imbalance. However, potential disadvantages are an extensive surgery and patients may not have these malalignments and/or soft tissue imbalances. One study of 42 patients found improvements in mean Knee Society function scores (40–58 points), mean Knee Society pain scores (15–47 points), and mean knee ROM (55–82°).¹² However, pain and function scores after revision TKA for arthrofibrosis are not as high as scores for revision TKA secondary to infection, instability, wear, and loosening.¹³

The ongoing debate about treatment options for patients undergoing revision TKA for arthrofibrosis warrants further investigation. The purpose of this study was to compare arthrolysis with polyethylene exchange vs. full component revision for arthrofibrosis following primary TKA. Specifically, we assessed: (1) re-operation rates at one and two years; (2) rates of surgical complications at one year; and (3) associated risk factors for requiring a lysis of adhesion.

2. Methods

2.1. Database selection

The “MKnee” subset of the Mariner database (PearlDiver Inc., Colorado Springs, Colorado) was reviewed. This database contains over 120 million Health Insurance Portability and Accountability Act-compliant records comprising commercial, Medicare, Medicaid, government, and cash payers. This is one of the largest nationwide databases that tracks patients in a longitudinal fashion, limiting potential type II errors. Patient populations and outcomes were identified using International Classification of Diseases (ICD) 9 and 10 procedural and diagnosis codes, Current Procedural Terminology (CPT) codes, patient demographics, prescriptions, and physician specialties. Institutional review board approval was waived for this publicly available database study.

2.2. Patient selection

All patients who underwent a primary TKA between January 1, 2010 and April 31, 2020 were identified based on ICD and CPT codes (n = 1,952,917). Patients who were diagnosed with arthrofibrosis following TKA were then identified using the following codes: ICD-9-D-71956, ICD-9-D-71856, ICD-10-D-M24669, ICD-10-D-M24662, ICD-10-D-M24661, ICD-10-D-M25661, ICD-10-D-M25662, ICD-10-D-M25669 (n = 219,239). Next, patients were characterized by either receiving a LOA via a partial component revision arthroplasty (CPT-27486) or receiving a full component revision arthroplasty (CPT-27487), n = 1290 and n = 1,120, respectively. The rate of arthrofibrosis within one year after primary TKA was 11.2%. Demographics and baseline characteristics were also defined by ICD codes and included age, sex, diabetes, obesity, morbid obesity, tobacco use, chronic obstructive pulmonary disease (COPD), fibromyalgia, gait abnormality, depression, alcohol abuse, anxiety, and hypertension. Post-operative outcomes and complications included re-operation at one and two years following either a partial or full component TKA revision. Prosthetic joint infections (PJIs), surgical site infections, and periprosthetic fracture rates were also assessed at one year. A subsequent analysis was conducted to determine associated risk factors for arthrofibrosis for patients requiring LOA. The base population of arthrofibrosis following TKA was examined via multivariate regressions. Within this population, those patients who had a specific comorbidity were compared to those who did not have the same comorbidity. The resulting odds ratio determined the increased, decreased, or non-significant risk for LOA.

2.3. Patient demographics

The mean age for both patient cohorts was 61 years (range, 40 to 84) (p = 0.958) (see Table 1). Women predominated in both cohorts (58 versus 42% and 62 versus 38.0%, p = 0.046) for polyethylene exchange and full revision, respectively. The full revision cohort had a higher incidence (46.5 versus 42.3%) of gait abnormalities, whereas, the polyethylene exchange cohort had a greater incidence of obesity (60.7 versus 56.6%). Additionally, all cohorts had similar baseline rates of diabetes, tobacco use, congestive heart failure (CHF), chronic obstructive pulmonary disease (COPD), fibromyalgia, depression, and alcohol abuse (all p ≥ 0.282).

Table 1.

Demographics and baseline characteristics.

Variables	Polyethylene Exchange Revision	Full Component Revision	p value
	(n = 1290) n (%)	(n = 1120) n (%)
Mean age in years (range)	61 (40–84)	61 (40–84)	0.958
Sex			0.046
Men	542 (42.0)	425 (38.0)
Women	748 (58.0)	695 (62.0)
Diabetes	617 (47.8)	545 (48.7)	0.714
Obesity	783 (60.7)	634 (56.6)	0.046
Tobacco Use	557 (43.2)	471 (42.1)	0.606
COPD	488 (37.8)	399 (35.6)	0.282
Fibromyalgia	240 (18.6)	213 (19.0)	0.836
Gait Abnormality	546 (42.3)	521 (46.5)	0.043
Depression	594 (46.1)	524 (46.8)	0.748
Alcohol Abuse	85 (6.6)	74 (6.6)	0.999
Anxiety	550 (42.6)	489 (43.7)	0.642
Hypertension	1091 (84.6)	950 (84.8)	0.911

SD: standard deviation; COPD: chronic obstructive pulmonary disease.

*Values < 11 hidden in accordance with PearlDiver confidentiality agreement.

2.4. Data analyses

Continuous variables including ages were compared using Student's t-tests. Categorical variables including demographics, comorbidities, re-operations, and complications were compared using chi-square tests. Multivariate regression analysses were performed to determine risk factors associated with PJIs, revisions, and the need for LOAs. This model included the following variables: age; sex; obesity as well as morbid obesity; diabetes; hypertension; CHF; COPD; depression; alcohol abuse; tobacco use; fibromyalgia; gat abnormality, and anxiety. All analyses were performed using R Studio (Statistics Department of the University of Auckland, New Zealand) with significance defined as p < 0.05.

3. Results

3.1. Re-operations

The difference in unadjusted incidence and odds of re-operation within one year following LOA in the form of polyethylene exchange TKA was non-significant compared to the full component revision (10.6 versus 12.9%, odds ratio (OR) 0.83, 95% confidence interval (CI) [0.64–1.06], p = 0.145). Similarly, at two years following revision the unadjusted comparison remained non-significantly different (14.3 versus 15.0%, OR 0.94 [0.75–1.18], p = 0.651). The adjusted models for re-operation within one and two years following LOA in the form of partial component revision was significantly higher than the full component revision cohort (OR 1.52 CI [1.07–2.17], p = 0.022 and OR 1.44 CI [1.06–1.97] p = 0.022) (see Table 2, Table 3, Table 4).

Table 2.

Postoperative outcomes and complication rates.

Variables	Polyethylene Exchange Revision		Full Component Revision		p-value
	(n = 1290)		(n = 1120)
	n	(%)	n	(%)
1-Year Revision Surgery	140	10.9	144	12.9	0.145
2-Year Revision Surgery	184	14.3	168	15.0	0.651
1-Year Complications
Periprosthetic Fractures	12	0.9	19	1.7	0.138
Prosthetic Joint Infections	57	4.4	45	4.0	0.700
Surgical Site Infections	59	4.6	58	5.2	0.552
Stiffness	169	13.1	143	12.8	0.856
Lysis of Adhesions	a		a		0.851
2-Year Complications
Periprosthetic Fractures	14	1.1	20	1.8	0.200
Prosthetic Joint Infections	65	5.0	47	4.2	0.377
Surgical Site Infections	65	5.0	64	5.7	0.520
Stiffness	186	14.4	158	14.1	0.873
Lysis of Adhesions	a		a		0.997

^aValues < 11 hidden in accordance with PearlDiver confidentiality agreement.

Table 3.

Odds Ratios, Outcomes and Complication Rates: Polyethylene Exchange vs Full Component Revision TKA.

Variables	OR (95% CI)
90-Day Manipulation under Anesthesia	0.44	0.27–0.74
1-Year Revision Surgery	0.83	0.64–1.06
2-Year Revision Surgery	0.94	0.75–1.18
1-Year Complications
Periprosthetic Fractures	0.54	(0.26–1.13)
Prosthetic Joint Infections	1.10	(0.74–1.65)
Surgical Site Infections	0.88	(0.61–1.27)
Stiffness	1.03	(0.81–1.31)
Lysis of Adhesions	0.65	(0.15–2.91)
2-Year Complications
Periprosthetic Fractures	0.60	(0.30–1.20)
Prosthetic Joint Infections	1.21	(0.83–1.78)
Surgical Site Infections	0.88	(0.61–1.25)
Stiffness	1.03	(0.82–1.29)
Lysis of Adhesions	1.16	(0.40–3.35)

OR: odds ratio; CI: confidence interval.

*Referent group: full component revision TKA.

Table 4.

Multivariate logistic regression for independent risk factors.

1-year all-cause revisions	ORa	95% CI	p-value
Male sex	1.47	1.31–1.65	<0.001
Age <60 years	1.03	0.71–1.43	0.888
Alcohol Abuse	1.76	1.10–2.67	0.012
Diabetes Mellitus	1.08	0.95–1.24	0.243
Obesity	0.95	0.83–1.08	0.436
Tobacco Use	1.04	0.89–1.21	0.590
Partial Cohort	1.52	1.07–2.17	0.022
2- year all-cause revisions
Male sex	1.37	1.23–1.52	<0.001
Age <60 years	1.14	0.84–1.52	0.381
Alcohol Abuse	1.79	1.18–2.60	0.003
Diabetes Mellitus	1.08	0.96–1.21	0.203
Obesity	0.94	0.83–1.06	0.322
Tobacco Use	1.00	0.87–1.15	0.958
Partial Cohort	1.44	1.06–1.97	0.022

OR: odds ratio; CI: confidence interval;

^aReferent group: full component revision TKA.

3.2. Surgery-related complications

The odds of peri-prosthetic fracture within one year (OR 0.54 [0.26–1.13], p = 0.138) and two years (OR 0.60 [0.30–1.20], p = 0.200) following partial component revision TKA was not different compared to full component revision. This was true for the unadjusted peri-prosthetic joint infections at one year (OR 1.10 [0.74–1.65], p = 0.700) and two years (OR 1.21 [0.83–1.78], p = 0.377) as well as surgical site infections at one year (OR 0.88 [0.61–1.27], p = 0.552) and two years (OR 0.88 [0.61–1.25], p = 0.520) (see Table 2, Table 3).

The overall adjusted risk for PJI following LOA in the form of partial component revision was significantly increased compared to the full component revision at one year (OR 1.93 CI [1.20–3.20], p = 0.008) and two years (OR 2.04 CI [1.30–3.30], p = 0.003) (see Table 5).

Table 5.

Multivariate logistic regression for independent risk factors.

1-year PJI	ORa	95% CI	p-value
Male sex	1.72	1.26–2.36	0.001
Age <60 years	0.82	0.29–1.84	0.671
Alcohol Abuse	2.69	0.92–6.23	0.038
Diabetes Mellitus	1.09	0.74–1.56	0.650
Obesity	0.72	0.48–1.06	0.112
Tobacco Use	1.00	0.65–1.48	0.986
Partial Cohort	1.93	1.20–3.20	0.008
2-year PJI
Male sex	1.88	1.39–2.55	<0.001
Age <60 years	0.89	0.34–1.88	0.779
Alcohol Abuse	2.99	1.13–6.54	0.013
Diabetes Mellitus	1.07	0.74–1.51	0.726
Obesity	0.71	0.48–1.03	0.085
Tobacco Use	0.96	0.63–1.42	0.858
Partial Cohort	2.04	1.30–3.30	0.003

OR: odds ratio; CI: confidence interval; PJI: prosthetic joint infection;

^aReferent group: full component revision TKA.

3.3. Associated risk factors for lysis of adhesions

A few factors reached statistical significance for patients who received a LOA (p < 0.05). These included age less than 60 years (OR 1.85 CI [1.67–2.03], p < 0.001), depression (OR 1.11 CI [1.02–1.21], p = 0.011), fibromyalgia (OR 1.23 CI [1.07–1.40], p = 0.003) and anxiety (OR 1.17 CI [1.07–1.29], p = 0.001) (See Table 6). Other assessed risk factors did not reach statistical significance (p > 0.05).

Table 6.

Multivariable regression model evaluating risk factors for requiring Lysis of Adhesions.

Variables	OR	95% CI	p value
Lysis of Adhesions
Age <60	1.85	(1.67–2.03)	<0.001
Male gender	1.03	(0.98–1.07)	0.260
Obesity	0.96	(0.90–1.03)	0.315
Diabetes	0.97	(0.91–1.02)	0.241
COPD	1.12	(1.03–1.21)	0.008
Depression	1.11	(1.02–1.21)	0.011
Alcohol Abuse	1.07	(0.81–1.39)	0.599
Tobacco Use	1.07	(0.98–1.17)	0.107
Fibromyalgia	1.23	(1.07–1.40)	0.003
Gait Abnormality	1.05	(0.91–1.20)	0.488
Anxiety	1.17	(1.07–1.29)	0.001

OR: odds ratio; CI: confidence interval; COPD: chronic obstructive pulmonary disease.

4. Discussion

Traditional management of recalcitrant arthrofibrosis following primary TKA primarily consists of polyethylene exchange or full component revision. However, concerns over re-operation rates after polyethylene exchange raises the question of the importance of a full component revision TKA.10, 11, 12 We sought to examine patients who develop arthrofibrosis and either received a arthrolysis with a polyethylene exchange or a full component revision TKA. We found higher risk of re-operation at one and two years following polyethylene exchange as compared to full revision.

This study is not without its limitations that are inherent with the use of a large database. However, an independent party performs an annual audit on the studied populations, thereby decreasing the chance of billing/coding errors. Examination of pertinent clinical factors such as knee range of motion, clinical outcome scores, and severity of arthrofibrosis could not be assessed due to the limitations of the database. However, identification of the study cohorts was performed methodologically, whereby, only patients who were actively tracked along the preset study period were included. Additionally, we could not assess surgery-specific variables such as surgical volume, rates of LOA versus revision TKA per surgeon, or intra-operative findings. Several factors, such as joint function score, patients' expectations, and patients’ physical condition may have an impact on the choices made by physicians and patients but were not assessed in this analysis. However, study cohorts were categorized based on specific procedure codes and timing of complications following the development of arthrofibrosis. This database is unique in its ability to track a large number of patients longitudinally over a decade of hospitalizations. While outcomes and complications 2 years after revision arthroplasty does not give the final verdict, it it provides an impetus for future studies. The strength of our study lies in the comparison of poly exchange versus full component revision utilizing large patient numbers that has not been examined in this context in previous literature.

Re-operations following polyethylene exchange or full component revision TKA may pose substantial morbidity to patients who have arthrofibrosis. The almost doubled one- and two-year re-operation rates in the polyethylene exchange TKA cohort compared to full component revision has been corroborated by multiple studies. In patients who required a manipulation under anesthesia (MUA), Werner et al.,¹³ demonstrated a subsequent revision rate of 4.8% compared to patients who did not require a MUA (2.0%). In a review of a Medicare and national database from 2005 to the first quarter of 2017, Cregar et al.,¹⁰ identified a cohort of LOA patients, 25.1% of whom underwent subsequent revision TKA within two years. This markedly higher rate of re-operation in comparison to the present study may be partially contributed by study population payer differences, with the aforementioned studies mostly comprised of Medicare recipients. Payer type is a known factor for patient outcomes. Starring et al.,¹⁴ assessed TKA outcomes in Medicare beneficiaries versus commercially insured. In the Medicare cohort, they found decreased ability to performs daily activities, worse physical function, and higher pain interference of physical, mental, and social activities. These deficiencies may also predispose patients to developing arthrofibrosis. Arthrofibrosis is a known risk factor for subsequent revision TKA. Our study further advances the current understanding that LOA in the form of polyethylene exchange revision TKA may not be the most effective operation in treating recalcitrant arthrofibrosis. Rather, considering a full component revision TKA may diminish the risk for further subsequent operations.

Other studies have suggested that polyethylene exchange revision TKA may lead to superior outcomes for patients who have arthrofibrosis.^15,16 Xiong et al., reported on the post-operative outcomes of 101 idiopathic stiff knees which underwent either isolated tibial insert exchange (n = 42) or component revision (n = 59).¹⁵ They found that the isolated tibial insert exchange cohort had lower revision rates than the component exchange group (16.7 versus 31.0%, p = 0.01). Additionally, greater rates of infection (5.1, 3 out of 59, versus 2.4%, 1 out of 42) and fracture (2.0, 1 out of 59, versus 0%, 0 out of 42) were shown for the full component revision cohort. However, their sample size was not as large as that of the present study and a total of 8 surgeons utilizing various implants performed the operations, which may cause increased variability. Keeney et al., in 2005 reviewed 23 patients (25 knees) who had limited approach, tibial insert downsizing, (n = 12) or comprehensive (n = 11) revision total knee arthroplasty for painful limited range of motion.¹⁶ They also demonstrated lower revision rates for limited approach, tibial insert downsizing, compared to comprehensive revision TKA (14.1 versus 25.9%).^15,16 While those results suggest that polyethylene exchange revision TKA may be the preferred surgical option for arthrofibrosis, potentially due to a less invasive procedure, there still needs to be further investigation to determine every outcome of treatment choices.¹⁵ The current study presents large up-to-date analyses of re-operation rates, surgical complication rates, and associated risk factors for requiring a LOA in regards to polyethylene exchange versus full component revision for TKA arthrofibrosis.

5. Conclusion

Recalcitrant arthrofibrosis managed with full component revision TKA was associated with markedly reduced one- and two-year re-operation rates as compared to polyethylene exchange. A subsequent analysis of risk factors for patients requiring a lysis of adhesions demonstrated that ages less than 75 years and fibromyalgia to be most predictive. These results are reflective of the recent increased interest towards performing full component revision TKA versus a polyethylene exchange revision for the treatment of arthrofibrosis.

Funding

None.

Data availability

Available in a respository upon request.

Ethical approval

IRB exemption due to retrospective nature and public database.

Authors’ contribution

ZC- Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Resources; Software; Supervision; Validation; Visualization; Roles/Writing - original draft; and Writing - review & editing. SB- Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Resources; Software; Supervision; Validation; Visualization; Roles/Writing - original draft; and Writing - review & editing. JD- Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Resources; Software; Supervision; Validation; Visualization; Roles/Writing - original draft; and Writing - review & editing. DH- Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Resources; Software; Roles/Writing - original draft; and Writing - review & editing. OS- Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Visualization; Roles/Writing - original draft; and Writing - review & editing. MM- Writing - review & editing. SP- Writing - review & editing. RD- Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Visualization; Roles/Writing - original draft; and Writing - review & editing. JN- Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Visualization; Roles/Writing - original draft; and Writing - review & editing. Use of AI tool- No use of AI tool.

Declaration of competing interest

ZC- None.

SB- None.

JD- None.

DH- None.

OS- None.

MM- None.

SP- None.

RD-is a paid consultant for Microport Orthopedics, receives Research support from a company or supplier as a Principal Investigator from Microport Orthopedics, Stryker, United Orthopedic Corporation, on the Medical/Orthopaedic publications editorial/governing board of Journal of Arthroplasty, Journal of the American Osteopathic Medicine Association, Orthopedic Knowledge, Journal of Knee Surgery, Knee and Board member/committee appointments for a society Of the arthritis foundation.

JN- receives research support from Johnson & Johnson, Biocomposites, CyMedica Orthopedics, Depuy Synthes Product, Flexion Therapeutics, Microport, Orthopedics, Orthofix, Patient-Centered Outcomes Research Institute (PCORI), Smith & Nephew, Stryker, Tissue Gene, United Orthopedic Corporation, is a Medical/Orthopaedic publications editorial/governing board for Journal of Knee Surgery and Board member/committee appointments for a society for Baltimore City Medical Society.