Serial ultrasonographic imaging can predict failure after meniscus allograft transplantation

Abstract

Introduction:

Treatment monitoring after meniscus allograft transplantation (MAT) is challenging. Ultrasonographic (US) imaging has been proposed as a modality that may allow for treatment monitoring after MAT, but has yet to be clinically validated for this purpose. The objective of this study was to assess the capabilities for serial US imaging during the first year after surgery to predict short-term MAT failure.

Methods:

Patients who had undergone Meniscus-only or Meniscus-Tibia MAT for treatment of medial or lateral meniscus deficiency were prospectively evaluated by US imaging at various time points after transplantation. Each meniscus was evaluated for abnormalities in echogenicity, shape, associated effusion, extrusion and extrusion with weightbearing (WB).

Results:

Data from 31 patients with a mean follow-up of 32 ± 16 (range, 12–55) months were analysed. MAT failure occurred in 6 patients (19.4%) at a median time point of 20 (range, 14–28) months with 4 (12.9%) converted to total knee arthroplasty. US imaging was effective for assessing MAT extrusion and imaging with WB demonstrated dynamic changes in MAT extrusion. US characteristics that were significantly associated with higher likelihood for MAT failure included abnormal echogenicity, localised effusion, extrusion with WB at 6 months, and localised effusion and extrusion with WB at 1 year.

Conclusions:

US assessments of meniscus allografts at 6 months after transplantation can effectively determine risk for short-term failure. Abnormal meniscus echogenicity, persistent localised effusion and extrusion with weightbearing were associated with 8–15 times higher odds for failure, which occurred at a median of 20 months post-transplantation.

Introduction

Symptomatic loss of functional meniscus affects millions of individuals annually, and although data regarding total health care costs are lacking, direct treatment costs, time lost from work, related socioeconomic impact and disability are significant.^1,2 If not effectively treated, meniscus deficiency inevitably leads to knee osteoarthritis (OA).^3–6 When indicated, meniscus allograft transplantation (MAT) can be a safe and effective treatment for meniscus deficiency with 5-year functional success rates from 75% to 90% reported.^7–10 Various types of allografts and surgical techniques have been used successfully for MAT.^7–13 Despite its overall success, MAT is associated with complications including graft tearing, extrusion, shrinkage and/or degeneration, arthrofibrosis, tibiofemoral articular cartilage deterioration and failure to restore knee function.^14,15 Reoperation rates as high as 32% have been reported with conversion to total knee arthroplasty (TKA) required in up to 19% of MAT cases.^14–17 Unfortunately, real-time treatment monitoring after MAT is challenging such that complications are often diagnosed relatively late in their development when they have already progressed to failure.

Treatment monitoring after MAT has primarily involved physical examination, radiographic assessments and patient-reported outcome measures.^8–11 Magnetic resonance imaging (MRI) has been employed for post-MAT treatment monitoring in research studies but has not been routinely implemented in the clinical setting due to associated costs, availability and limitations.^18–21 Ultrasonographic (US) imaging has been proposed as a modality that may allow for readily available cost-effective treatment monitoring after MAT but has yet to be clinically validated for this purpose.^21–23 Therefore, the objective of the present study was to assess the capabilities for serial US imaging during the first year after surgery to predict short-term MAT failure. The study was designed to test the hypothesis that previously defined US imaging characteristics of meniscal health would be significantly associated with short-term MAT failure.

Methods

Patients were prospectively enrolled in an institutional review board (IRB)-approved registry (NCT02503228, NCT03257371) designed to follow lifelong outcomes after osteochondral and/or meniscus allograft transplantation at the authors’ institution. Patients were included for analyses when they had undergone MAT for treatment of medial or lateral meniscus deficiency from 2017 through 2021 and had consented to participate in an IRB-approved (#2008415) prospective assessment of US imaging with at least 1 year of follow-up after MAT. Patient demographics, prior surgeries, MAT surgery data, complications, reoperations and failures were documented. Demographic and operative data including age, sex, body mass index (BMI), previous surgeries, MAT type and concurrent procedures were collected from the electronic medical record (EMR). Patients with incomplete data or insufficient follow-up were excluded.

MAT surgeries were performed on patients with symptomatic deficiency of one meniscus in one knee who chose this treatment option over other non-surgical or surgical options as indicated and were approved for coverage by their insurance provider. Radiographically size-matched fresh (viable) meniscus and osteochondral allografts preserved using the Missouri Osteochondral Preservation System (MOPS®) were obtained from an AATB-accredited tissue bank (MTF Biologics, Edison, NJ, USA) and used within 56 days after recovery in conformance of the tissue to the US Food and Drug Administration (FDA) classification of a Human Cell and Tissue Product under Section 361 of the Public Health Services Act. MAT was performed using one of two transplantation techniques selected based on the nature of the defects being treated:

• Meniscus – Meniscus allografts were transplanted using a double bone plug technique with suspensory fixation and meniscotibial ligament reconstruction, as previously described. ¹⁰ Meniscocapsular sutures were placed at the surgeon’s discretion.

• Meniscus-Tibia – Meniscus-tibial-osteochondral allografts were transplanted using a custom-cut tab-in-slot technique with preservation of the meniscal roots and meniscotibial (coronal) ligament and fixation using bioabsorbable implants, as previously described.^11,13 Meniscocapsular sutures were placed at the surgeon’s discretion.

If necessary, procedures to address symptomatic comorbidities in the same knee including ligament reconstruction, chondroplasty and/or osteochondral allograft (OCA) transplantation were performed concurrently.

Prescribed postoperative rehabilitation consisted of 2 weeks without weightbearing followed by 4 weeks of weightbearing as tolerated, all with a hinged knee brace locked in extension.^10,11 Quadriceps strengthening started immediately postoperatively and included passive knee range of motion for the first 6 weeks. Patients were allowed to progress to full weightbearing at 6–7 weeks postoperatively, and knee flexion was also limited to 90° for the first 6 weeks. Jogging was typically allowed at 5 months postoperatively at the earliest and return to sports and unrestricted activities after 8–10 months with physician’s clearance. The protocol was adjusted based on concomitant surgeries and assessment of patient progress by the physician and physical therapist.

Based on scheduling of standard-of-care clinical follow-up appointments, patients were evaluated by US imaging of the treated knee at 2 weeks, 6 weeks, 6 months and 1 year after MAT using a portable ultrasound machine (GE Logiq e with 12L-RS 5-13 MHz Wideband Linear probe, GE Health Care, Milwaukee, WI, USA). US examination was performed by one board-certified veterinary radiologist (CRC) with more than 25 years’ experience in canine, equine and human musculoskeletal ultrasonography including development and validation of meniscal assessments, who was blinded to treatment type and time point. Each patient was initially placed supine with the knee extended, so that the anterior horns of the menisci could be examined. The patient was then placed prone so the posterior horns of each meniscus could be examined. Each meniscus was evaluated for abnormalities in echogenicity, shape, associated effusion and extrusion, as follows:^22,23

• Abnormal echogenicity – meniscus parenchyma not uniformly hyperechoic with uniform echotexture

• Abnormal shape – cross section of meniscus is not a sharp, well-defined triangle that parallels the apposing articular surfaces of the femoral and tibial condyles

• Abnormal effusion – any localised fluid associated with the meniscus that is more than a 1–2 mm uniform anechoic line interposed between superior and inferior margins of meniscus and femoral and tibial surfaces, respectively, or any fluid lines or pockets abaxial to the peripheral margin of the meniscus

• Abnormal displacement (extrusion) – peripheral meniscus margin extends abaxial to the cortical margins of femur and tibia

In addition, meniscus extrusion with weightbearing was assessed beginning 6 weeks post-MAT, as allowed, by capturing a longitudinal-axis (coronal cross-section) image of the mid-body and posterior horn prior to and after the patient placed equally distributed weight on both lower extremities while standing with the knee in extension. The peripheral edge of the meniscus allograft and cortical margin of tibia were defined in each image and the machine’s calibrated measuring tool was used to determine meniscal offset (mm) of the edge of meniscus relative to the subjacent tibial cortex. If weightbearing offset was ⩾2 mm more than non-weightbearing offset, extrusion with weightbearing was recorded as abnormal.^24,25

All reported complications and reoperations were recorded in the EMR. For the purposes of the present study, MAT failure was defined as a second operation to revise the meniscus allograft or for conversion to TKA for any reason. The decision to pursue revision surgery or TKA was based on the attending surgeon’s discussion of joint pathology, treatment options and related prognosis in conjunction with patient preference and informed consent.

Cases were included for statistical analyses when complete data were available for patients undergoing MAT for the first time (primary transplant) with at least 1 year of follow-up. Descriptive statistics were calculated to report means, standard deviations, medians, ranges and percentages. Data were normally distributed such that groups were compared using unpaired t tests (age, BMI, previous surgeries, follow-up duration) or two-tailed Fisher’s exact tests for proportions (sex, laterality, outcome, US characteristics). When statistically significant differences in proportions were noted, odds ratios were calculated. All analyses were performed using SigmaStat 4.0 (Systat, San Jose, CA, USA). Statistical significance was set a priori at p < 0.05.

Results

From a total of 93 potentially eligible patients undergoing MAT during the study period, 38 patients consented and enrolled and 31 patients (Meniscus: n = 12; Meniscus-Tibia: n = 19) completed the study with a mean clinical follow-up of 32 ± 16 (range, 12–55) months (Figure 1).

Figure 1.

Flowchart for number of patients included in the study.

No statistically significant differences between cohorts were noted for age (p = 0.15), proportion of males versus females (p = 1), BMI (p = 0.37), proportion of medial versus lateral MAT (p = 0.42), or follow-up duration (p = 0.33). Patients in both MAT cohorts had complex knee problems based on history of previous surgeries and comorbidities. All patients had undergone at least one previous surgery for the affected knee with means of 2.1 and 2.4 previous surgeries per patient prior to Meniscus or Meniscus-Tibia transplants, respectively. Number of prior surgeries was not significantly different between cohorts (p = 0.6) (Table 1).

Table 1.

Patient demographic, surgical and outcomes data.

	Meniscus	Meniscus-Tibia
Age, years	32.8 ± 12.6	38.8 ± 9.9
Male:Female	6:6	10:9
BMI, kg/m2	27.8 ± 4.9	29.7 ± 5.7
Medial:Lateral	10:2	12:7
Previous surgeries	2.1	2.4
Follow-up, months	29.4 ± 14	37.1 ± 15
Successful:Failed	10:2	15:4

BMI: body mass index.

Concurrent procedures were performed to treat comorbidities in the knee undergoing MAT in 8 Meniscus patients (67%) and 17 Meniscus-Tibia patients (89%) (p = 0.17). Concurrent procedures performed for Meniscus patients included ACL reconstruction (n = 5), OCA transplantation of the corresponding femoral condyle (n = 3), and posterior medial corner ligament reconstruction (n = 1). Concurrent procedures performed for Meniscus-Tibia patients included ACL reconstruction (n = 2) and OCA transplantation of the corresponding femoral condyle, patella, and/or trochlea (n = 17).

Based on serial US imaging assessments, MAT shape was maintained in the vast majority of cases throughout the 1-year study period. Echogenicity of MATs varied based on time point and patient. Localised effusion persisted in nearly all cases through 6 weeks post-MAT and progressively resolved in the majority of successful MATs by 1 year after transplantation. US imaging was effective for assessing MAT extrusion and imaging with weightbearing demonstrated dynamic changes in MAT extrusion (Figures 2 and 3).

Figure 2.

Representative US images from patients in this study at 6 months after transplantation showing. (a) Successful lateral Meniscus-Tibia MAT with no US abnormalities during weightbearing. (b) Successful medial Meniscus MAT with no US abnormalities during weightbearing. (c) Subsequently failed lateral Meniscus-Tibia MAT showing abnormal echogenicity, shape, and displacement (extrusion) during weightbearing. (d) Subsequently failed medial Meniscus MAT showing abnormal echogenicity, localised effusion, and displacement (extrusion) during weightbearing. (e, f) Subsequently failed medial Meniscus MAT showing non-weightbearing extrusion (e) and increased extrusion with weightbearing (f).

Figure 3.

Serial US images at (a) 2 weeks, (b) 6 weeks, (c) 6 months and (d) 1 year showing progression of abnormal echogenicity, shape and displacement (extrusion) for a patient in this study who subsequently failed medial Meniscus-Tibia MAT 16 months after transplantation.

MAT failure occurred in 6 patients (19.4%; Meniscus (2 medial) (16.7%), Meniscus-Tibia (3 medial, 1 lateral) (21.1%), p = 1). For MAT failures (4 male, 2 female), mean age was 38 ± 9.3 years, mean BMI was 28 ± 2.9, and median follow-up was 36 (range, 19–32) months. Failures occurred at a median time point of 20 (range, 14–28) months after transplantation surgery. For this patient population, age, sex, BMI, number of previous surgeries, laterality of MAT, number of concurrent procedures, and type of MAT were not significantly associated with higher likelihood for failure (p > 0.3). US characteristics that were significantly associated with higher likelihood for MAT failure included abnormal echogenicity at 6 months (p = 0.038; OR = 11.5), localised effusion at 6 months (p = 0.022; OR = 12.9), extrusion with WB at 6 months (p = 0.043; OR = 8), localised effusion at 1 year (p = 0.014; OR = 14.7), and extrusion with WB at 1 year (p = 0.014; OR = 14.7) (Table 2).

Table 2.

Percentage (and number) for abnormal ultrasonographic findings for successful (n = 25) versus failed (n = 6) meniscus allograft transplants.

Cohort	2 weeks
	Echogenicity	Shape	Effusion	Extrusion	Extrusion with WB
Successful	44% (11)	0% (0)	96%(24)	12% (3)	na
Failed	50% (3)	0% (0)	83% (5)	17% (1)	na
	6 weeks
	Echogenicity	Shape	Effusion	Extrusion	Extrusion with WB
Successful	32% (8)	0% (0)	92% (23)	20% (5)	24% (6)
Failed	33% (2)	0% (0)	83% (5)	17% (1)	33% (2)
	6 months
	Echogenicity	Shape	Effusion	Extrusion	Extrusion with WB
Successful	8% (2)*	8% (2)	28% (7)*	20% (5)	20% (5)*
Failed	50% (3)*	17% (1)	83% (5)*	17% (1)	67% (4)*
	1 year
	Echogenicity	Shape	Effusion	Extrusion	Extrusion with WB
Successful	24% (6)	4% (1)	12% (3)*	16% (4)	12% (3)*
Failed	33% (2)	17% (1)	67% (4)*	33% (2)	67% (4)*

WB: weightbearing; na: not applicable as patients were restricted from weightbearing for 2 weeks following MAT.

^*Statistically significant differences in proportions.

Revision of medial MAT was performed in two patients (24-year-old female, 37-year-old male), one in each cohort, such that overall success rate for this patient population was 87.1%. The other four MAT failures were converted to TKA.

Discussion

The results of the present study allow for acceptance of the hypothesis in that abnormalities in defined US imaging characteristics of meniscal health were significantly associated with short-term meniscus allograft transplant failures. In the patient population studied, reoperation rate was 19% with conversion to TKA required in 13% of cases, which fall into the lower ranges reported in previous studies^8–11,14,17. Abnormal echogenicity at 6 months, localised effusion at 6 months and 1 year, and extrusion with weightbearing at 6 months and 1 year were associated with a significantly higher likelihood for MAT failures, which occurred at a median of 20 months after transplantation. Age, sex, BMI, number of previous surgeries, laterality of MAT, number of concurrent procedures, and type of MAT were not significantly associated with higher risk for MAT failure in the patient population studied.

A readily-available cost-effective method for treatment monitoring after MAT that allows for accurate identification of impending complications prior to failure is highly desirable. While prior efforts have identified biomarkers associated with meniscal pathology, there are currently no real-time methods for predicting graft survivorship success or failure following MAT.^{9,10,13–15,18,21,26–28} Among the clinically applicable diagnostic imaging modalities, MRI has been most commonly used for assessment of meniscus health and disease.^18–20 MRI characteristics of progressive MAT healing, as well as features of failure, have been well described.^18–20 While MRI provides robust sensitivity and specificity for diagnosis of meniscal pathology, it is not ideal for serial post-MAT treatment monitoring based on associated costs, availability, and lack of ease in allowing for dynamic weightbearing or functional assessments.^18–22 Comparative imaging using an upright MRI would be highly desirable for future validation studies. ²⁹

US imaging has been used clinically for assessment of meniscus health and disease, but has not previously been employed for post-MAT treatment monitoring to the authors’ knowledge.^19–23 In the present study, US findings related to meniscal echogenicity, localised effusion, and extrusion with weightbearing at 6 months and/or 1 year after MAT were associated with 8–15 times higher odds for failure at a median of 20 months after transplantation. Interestingly, changes in shape of transplanted menisci were uncommon in the present study, which could be related to the use of fresh (viable) allografts that may mitigate tissue necrosis, degeneration, and shrinkage.^10,11,13,21 The lack of uniformly hyperechoic echotexture noted at 6 months in MATs that went on to fail is likely associated with failure of allograft revascularization as well as disruption and degradation of the extracellular matrix.^19–22,26 While localised effusion was commonly noted for successful and failed MATs through at least 6 weeks after transplantation, persistent meniscus-associated effusion was indicative of impending failure. This persistent localised effusion may be caused by biologic and/or biomechanical abnormalities associated with allograft integrity, composition, integration, and/or function.^19–22,26 Importantly, non-weightbearing extrusion of MATs through 1 year after transplantation did not differentiate subsequent success from failure, while extrusion with weightbearing at 6 months and 1 year post-MAT was highly associated with failure. Similar to MRI, ¹⁸ non-weightbearing MAT extrusion did not progress over serial assessments during the 1-year postoperative period while MAT extrusion with weightbearing did progress for MATs that subsequently failed. This finding suggests that static, non-weightbearing imaging may be associated with a high ‘false negative’ rate for diagnosis of MAT extrusion and highlights a key advantage for the use of US imaging for treatment monitoring after MAT as it allows for dynamic (functional) assessment as soon as weightbearing is allowed. Taken together, these data suggest that from the US assessments and time points evaluated, meniscus extrusion with weightbearing at 6 months and 1 year after MAT provides the most effective prediction of clinical failure. While additional post-transplantation time points for US assessment of MAT may be valuable, all of the documented failures in the present study could be predicted by 6-month and/or 1-year sonographic evaluations.

Limitations to the present study must be considered when interpreting and applying the results. The number and type of patients that completed the study render the data subject to selection bias. Based on the principles of uncoerced participation and penalty-free withdrawal from human subjects research, only one-third of eligible MAT patients enrolled in and completed the study. This relatively small number of patients with complex knee pathology may have resulted in type II errors for differences that were not statistically significant, and also limits the generalisability of the data. Similarly, the use of only a single highly experienced ultrasonographer performing assessments that are substantially user-dependent limits the broad applicability of findings. In addition, the lack of any mid- or long-term data limits application of the results to only short-term outcomes in patients undergoing fresh meniscus allograft or meniscus and tibial osteochondral allograft transplantation for complex knee disorders.

Conclusion

Taken together, the results of the present study suggest that US assessments of meniscus allografts at 6 months after transplantation can effectively determine risk for short-term failure. Abnormal meniscus echogenicity, persistent localised effusion, and extrusion with weightbearing were associated with 8–15 times higher odds for failure at a median of 20 months after MAT. Based on timing for intervention, differentiating capabilities (odds ratio), and relationship to meniscus function, extrusion with weightbearing at 6 months after MAT was considered to be the most clinically useful sonographic assessment for predicting clinical success versus failure. Identifying high risk for failure at this early postoperative time point may allow for implementation of management strategies to mitigate treatment failure. Multicenter studies that include a larger and more diverse patient population are needed to validate these findings prior to broad application to routine clinical use.