Abstract
Background:
Focal chondral defects in the knee lead to swelling, discomfort, and mechanical symptoms, resulting in disability and loss of function. Matrix-induced autologous chondrocyte implantation (MACI) has emerged as a viable treatment option, with satisfactory clinical outcomes reported compared with microfracture. Patient and lesion-specific variables associated with MACI after biopsy remain largely underreported.
Purpose:
To examine patient- and defect-specific characteristics associated with conversion to MACI after biopsy from a single institution.
Study Design:
Cohort study; Level of evidence, 3.
Methods:
A retrospective chart review of prospectively collected data from June 2018 to March 2025 was performed. Inclusion criteria included patients ≥15 years of age with an arthroscopically confirmed focal International Cartilage Regeneration & Joint Preservation Society (ICRS) grade 2 to 4 knee chondral lesion undergoing chondral biopsy (phase 1). Patient and chondral defect characteristics were assessed using univariate and multivariate regression analyses to determine factors associated with conversion to MACI (phase 2).
Results:
A total of 104 patients (mean patient age, 31.4 ± 11.4 years) undergoing chondral biopsy were identified. MACI was performed in 28.8% (30/104) of patients at a mean of 9.8 ± 10.2 months from biopsy. Patients with a history of previous knee surgery (P = .003), patellar lesions (P = .014), and ICRS chondral defect grade ≥3 (P = .048) had a higher probability of undergoing MACI. Patients undergoing concomitant procedures during biopsy (P < .001), as well as patients with preoperative mechanical symptoms (P = .014), were less likely to undergo MACI. Multivariate logistic regression found that patellar lesions (OR, 18.85; 95% CI, 1.70-208.50; P = .017) were independently associated with a greater likelihood of conversion to MACI, while concomitant procedures during biopsy (OR, 0.22; 95% CI, 0.06-0.81; P = .022) were associated with a significantly lower likelihood of implantation.
Conclusion:
MACI was performed in 28.8% of patients undergoing biopsy for symptomatic chondral lesions. Patients with patellar defects exhibited a significantly higher likelihood of MACI, with increased conversion rates associated with previous ipsilateral knee surgery and/or more advanced chondral defects. Patients undergoing concomitant procedures at biopsy were significantly less likely to require MACI, with lower conversion rates associated with a history of preoperative mechanical symptoms.
Focal, full-thickness chondral defects in the knee are common 11 and often lead to pain secondary to loss of the chondroprotection to the underlying bone, resulting in subchondral edema, effusion, and potentially mechanical symptoms, leading to disability and loss of function.6,12,25 Despite their ability to resist loading and shear forces, native chondrocytes lack regenerative potential, and chondral defects may expand by as much as 0.11 cm2 per month once identified on diagnostic arthroscopy.3,21,34,35 Although focal chondral defects are present in up to 66% of patients undergoing knee arthroscopy, treating symptomatic chondral defects remains challenging due to variable lesion location, size, and depth.36,37,41,43
Matrix-induced autologous chondrocyte implantation (MACI) has emerged as the third-generation chondral restoration technique using a porcine type 1/type 3 collagen bilayer, seeded with chondrocytes, fixated within the prepared chondral defect using fibrin glue. 1 MACI is performed in 2 phases, with phase 1 consisting of diagnostic arthroscopy to identify the size, location, and depth of the chondral defect, with debridement of any unstable chondral tissue and, in appropriate cases, procurement of a chondral biopsy specimen. Phase 2 includes implantation of the cultured cells within the chondral defect, generally using an open approach. In patients reporting symptomatic improvement after phase 1, phase 2 implantation may not be performed. The need for and timing of phase 2 vary and depend on symptom resolution, emphasizing the need for patient-specific treatment planning.
It remains largely unknown which patients are more likely to proceed to phase 2 MACI implantation, especially in those patients who become asymptomatic after the phase 1 biopsy. This may lead to a dilemma when counseling patients on the need for the second (phase 2) surgery for the treatment of their chondral defect. The purpose of this investigation was to determine variables predictive of progression from the phase 1 chondral biopsy to phase 2 MACI implantation. Specifically, we sought to examine patient- and defect-specific characteristics associated with conversion to MACI after biopsy. We hypothesized that larger defect size, patellar lesions, high-grade lesions, and previous ipsilateral knee surgery would be associated with an increased likelihood of progression to MACI implantation.
Methods
Patient Population
Before study initiation, institutional review board approval was obtained from Washington University to conduct a retrospective review of a prospectively collected database of patients undergoing phase 1 MACI chondral biopsy and/or conversion to phase 2 MACI implantation between June 2018 and March 2025. All procedures were performed by 1 of 5 fellowship-trained sports medicine surgeons (R.H.B., D.M.K., M.J.M., M.V.S.) at a single quaternary academic teaching institution, with a minimum 6-month follow-up from the time of cartilage biopsy. Patients were treated without the intention to progress to MACI implantation and were included regardless of the presence of concomitant procedures performed at the time of biopsy or implantation. Additional inclusion criteria included patients aged ≥15 year, and an International Cartilage Regeneration & Joint Preservation Society (ICRS) grade 2 to 4 chondral lesion confirmed arthroscopically during biopsy (Figure 1). Patients were excluded from consideration for MACI biopsy if they possessed degenerative changes within the affected compartments not amenable to MACI (eg, advanced/diffuse degenerative changes) or if they had underlying inflammatory arthritis. Patients who underwent chondral fragment repair or microfracture were also excluded, as violation of the underlying subchondral bone has been reported to yield inferior outcomes after MACI.28,33,39
Overview of study inclusion/exclusion criteria. ICRS, International Cartilage Regeneration & Joint Preservation Society; MACI, matrix-induced chondrocyte implantation.
Data Collection
For eligible patients meeting the inclusion criteria, a retrospective chart review was performed with the following variables collected: patient age, sex, body mass index (BMI), knee laterality (right vs left), mechanical axis knee alignment (varus, valgus [±3° deviation from neutral hip-knee-ankle angle], or neutral) and patellofemoral alignment as determined from standard radiographs (anteroposterior, lateral, and axial [Merchant or sunrise] views), biopsy method (loose body removal vs chondral harvest), patient symptoms (eg, knee pain, swelling, and mechanical pain), history of previous trauma, workers’ compensation status, defect location (ie, patella, medial condyle, lateral condyle, trochlea, or tibial plateau), number of defects (ie, single vs multiple), defect size (length vs width; cm2), defect depth (mm), involvement of subchondral bone, ICRS grade, previous surgical procedures, and any concomitant procedures performed at the time of biopsy and/or implantation. Concomitant procedures included anterior cruciate ligament reconstruction, loose-body removal, medial patellofemoral ligament reconstruction, lateral retinacular release, meniscectomy, meniscal repair, and/or medial retinacular imbrication.
Preoperative Evaluation
Patients underwent a thorough preoperative evaluation, including individual patient history with reported accounts of traumatic knee injury, symptom characteristics (eg, pain severity, location, mechanical symptoms, instability, and activating and alleviating factors), symptom duration, and any reported history of nonoperative treatment (ie, oral medications, supervised physical therapy, corticosteroid injection, etc). Preoperative physical examination included assessment of knee effusion, localized tenderness, active and passive range of motion (ROM), the McMurray test, Lachman grading, and anterior and posterior drawer tests. Preoperative imaging evaluation included bilateral anteroposterior standing, Rosenberg, lateral, and Merchant view knee radiographs, and magnetic resonance imaging to evaluate the integrity of the chondral surfaces, ligamentous structures, and soft tissues.
Biopsy Technique
One of 5 fellowship-trained sports medicine surgeons performed all phases of the biopsy procedure. Before the operation, patients were evaluated under general anesthesia for knee ROM, alignment, and stability. Diagnostic arthroscopy was performed with the patient in the supine position to assess the chondral integrity of all knee compartments and to evaluate for any loose bodies and/or synovitis. The location, number, size, and grade of chondral defects were recorded. Debridement and chondroplasty were performed for all cartilage lesions, in addition to the removal of any loose bodies. Final chondral defect measurements were obtained using a standard arthroscopic probe, and in patients deemed eligible for chondral restoration, a chondral biopsy specimen measuring approximately 5 × 8 mm was obtained from a nonweightbearing aspect of the joint, typically the lateral intercondylar notch, or loose chondral body if the loose fragment was noted to contain healthy-appearing (ie, smooth and firm, without evidence of softening or fissuring) cartilage. 40 Any necessary concomitant procedures to the menisci or collateral/cruciate ligaments were then performed. Chondral biopsy specimens were placed in a proprietary medium and sent to Vericel for storage.
During the postoperative period, in patients with persistent symptoms, concordant with known chondral pathology appreciated at the time of biopsy, discussion was held with the patient regarding the potential for chondrocyte implantation using the MACI technique. Generally, at least 6 weeks was allowed for chondrocyte processing to be completed, during which time the patient’s symptoms were closely monitored. Patients were informed that the chondral biopsy specimen could only be utilized at a maximum of 5 years from tissue acquisition.
Rehabilitation Protocol and Postoperative Management
Postoperative rehabilitation after MACI biopsy included a regimented course of supervised physical therapy focused on regaining knee ROM, progressing to strengthening exercises, and gradual progression to weightbearing as tolerated, with additional considerations for patients undergoing concomitant procedures. Passive ROM exercises (ie, manual and/or continuous passive motion) were initiated immediately, with full active ROM as tolerated after 4 weeks in patients with tibiofemoral defects and after 6 to 10 weeks in patients with patellofemoral defects. Full weightbearing was allowed immediately for patients withisolated/low-grade patellar lesions, after 5 to 6 weeks for patients with multiple/high-grade patellofemoral lesions, and after 7 to 9 weeks for patients with tibiofemoral lesions. Bracing was generally utilized until full weightbearing was comfortably reestablished. Postimplantation recovery protocols were determined at the discretion of the operating surgeon and followed established consensusguidelines, as well as those recommended by Vericel.10,13,16,27
Data Analysis
Deidentified patient characteristics were compiled and analyzed using Microsoft Excel (Version 16.96). Variables including patient age, sex, BMI, knee laterality, knee alignment, biopsy method, previous surgical procedures, patient symptoms, history of previous knee trauma, defect location, number of defects, defect area (cm2), and ICRS defect grade were calculated as mean with standard deviation. A comparative analysis of categorical data was performed using chi-square tests, while independent t tests were used to analyze and compare continuous data. Multivariate logistic regression using SPSS (Version 29.0.2.0; IBM) was conducted to evaluate the predictability of categorical and continuous variables for conversion from MACI biopsy to implantation. For multivariate analysis, all significant variables from univariate analysis were entered into a backward regression, as the limited sample size (n = 104) precluded inclusion of all factors in a single model. Logistic regression was run on the final 4 variables frombackward regression. Data are presented with 95% confidence intervals, with statistical significance set at a P value <.05.
Results
Overall Patient Characteristics
A total of 104 patients who underwent MACI biopsy during the study period were identified. The mean patient age at the time of biopsy was 31.4 ± 11.4 years (range, 15.0-58.5 years), with 50% (52/104) of patients being male (Table 1). The mean BMI at the time of biopsy was 27.7 ± 5.6 kg/m2 (range, 18.5-42.3 kg/m2), and the mean overall chondral defect area was 4.6 ± 3.9 cm2 (range, 0.01-27.5 cm2), with 66.3% (69/104) of patients possessing a single chondral lesion, and 33.7% (35/104) possessing ≥2 lesions. The most common defect location was the patella (49.3% [73/148]), followed by the medial femoral condyle (MFC) (20.3% [30/148]) and the trochlea (18.9% [28/148]). The mean ICRS chondral defect grade was 3.1 ± 0.6 (range, 2.0-4.0) (Table 2).
Patient Characteristics a
Data are presented as n (%) or mean ± SD. P values in bold represent statistical significance (P < .05). BMI, body mass index; MACI, matrix-induced autologous chondrocyte implantation.
Chondral Defect Characteristics a
Data are presented as n (%) or mean ± SD. P values in bold represent statistical significance (P < .05). ICRS, International Cartilage Regeneration & Joint Preservation Society; MACI, matrix-induced autologous chondrocyte implantation.
Implantation Patient Characteristics
Overall, 29% (30/104) of patients progressed to MACI implantation after biopsy. The mean time to implantation was 9.8 ± 10.2 months (range, 1.5-52.4 months) from biopsy. The mean patient age at implantation was 33.7 ± 11.8 years (range, 15.0-58.0 years), with 36.7% (11/30) of patients being male (Table 1). The mean patient BMI at implantation was 28.3 ± 5.3 kg/m2 (range, 20.6-39.5 kg/m2), while the mean overall chondral defect area was 3.2 ± 3.1 cm2 (range, 0.60-15.0 cm2), with 70.0% (21/30) of patients possessing a single chondral lesion. The most common chondral defect location in patients undergoing implantation was the patella (65.1% [28/43]), followed by the MFC (14.0% [6/43]) and the trochlea (14.0% [6/43]). The mean ICRS chondral defect grade was 3.2 ± 0.5 (range, 2.5-4.0) (Table 2).
Univariate Analysis
Univariate analysis revealed that a significantly greater proportion of patients who had undergone knee surgery before MACI biopsy (23.3% [7/30] vs 4.1% [3/74]; P = .003) and who had a chondral lesion of ICRS grade ≥3 (95.7% [22/23] vs 77.3% [51/66]; P = .048) went on to MACI implantation. When comparing the performance of MACI implantation versus biopsy alone based on defect location, 65.1% (28/43) of patients with patellar chondral defects underwent implantation (P = .014). Patients undergoing concomitant procedures at the time of MACI biopsy (59.5% [44/74] vs 23.3% [7/30]; P < .001), including concomitant soft tissue (40.5% [30/74] vs 13.3% [4/30]; P = .007) (Table 3) and meniscal (27.0% [20/74] vs 6.7% [2/30]; P = .021) procedures, were significantly less likely to require MACI implantation. Patients reporting mechanical symptoms were also significantly less likely to undergo implantation (85.1% [63/74] vs 63.3% [19/30]; P = .014). No significant differences between groups were appreciated regarding the location of the chondral defect, number of chondral defects, individual chondral defect size, or total chondral defect area. No statistically significant differences were observed regarding patient sex, age, or BMI.
Concomitant Procedures Performed During Chondral Biopsy a
Data are presented as n (%). ACLR, anterior cruciate ligament reconstruction; MACI, matrix-induced autologous chondrocyte implantation; MPFLR, medial patellofemoral ligament reconstruction.
Multivariate Analysis
Backward multivariate analysis was performed using 7 significant variables from the univariate analysis, including previous ipsilateral knee surgery, concomitant procedure during biopsy, meniscal procedure during biopsy, soft tissue repair procedure during biopsy, presence of mechanical symptoms, presence of a patellar lesion, and ICRS grade ≥3 chondral defect. Among these variables, previous ipsilateral knee surgery, concomitant procedures performed during biopsy, presence of a patellar lesion, and ICRS grade ≥3 chondral lesions were selected for a multivariate logistic regression model to assess their association with the progression to implantation (R2= 0.444; P < .001). Progression to implantation surgery was sign.0ificantly more likely in patients with patellar lesions (OR, 18.85; 95% CI, 1.70-208.50; P = .017). Patients were less likely to proceed to implantation if they underwent a concomitant procedure during biopsy (OR, 0.22; 95% CI, 0.06-0.81; P = .022).
Discussion
The primary findings of this retrospective study were that 29% of patients undergoing MACI biopsy progressed to MACI implantation. Patients with patellar defects exhibited a significantly greater likelihood of implantation (OR, 18.85), while those undergoing concomitant procedures at biopsy were significantly less likely to require implantation (OR, 0.22). Patients with a history of previous ipsilateral knee surgery and/or ICRS grade ≥3 chondral defects had a significantly greater incidence of undergoing implantation, while the presence of preoperative mechanical symptoms was associated with a decreased incidence of progression to implantation.
The observed implantation rate after biopsy of 29% aligns closely with previous investigations. Pasic etal 32 conducted a retrospective case series of 46 patients who underwent chondral biopsy to investigate the conversion rate to MACI/autologous chondrocyte implantation (ACI) for focal chondral defects. They reported an overall transplantation rate (including MACI/ACI and osteochondral allograft transplantation) of 26%. The authors also reported no association between lesion size or lesion location on progression to phase 2 MACI implantation, while no other risk factors for implantation were appreciated. Mason etal 26 conducted a retrospective cohort study including 71 patients who underwent phase 1 chondral biopsy and found a 35% rate of progression to phase 2 MACI implantation. They identified larger defect size (mean, 5.2 cm2 in the implantation group vs 3.3 cm2 in the biopsy-only group) and older age (≥26 years) as specific predictors for progression. While Mason etal also identified persistent symptoms after phase 1 as the indication for progression to phase 2, the differences in identified predictors may be secondary to methodological variations, such as differences in sample size (104 vs 71 patients), with a larger sample size providing greater statistical power to detect alternative associations. Moreover, our extended period of data collection of 2 years (6.8 vs 4.8 years) likely captures a broader range of patient outcomes compared with Mason etal.
Patients with previous ipsilateral knee surgery demonstrated a significantly increased likelihood of implantation. A history of previous knee surgery may indicate that initial symptoms were not fully resolved during previous surgical intervention, potentially indicating a more severe or persistent chondral lesion that is less amenable to isolated debridement. A retrospective study of 126 patients from the German Cartilage Registry by Weißenberger etal 44 found that arthroscopic debridement for isolated focal chondral defects significantly improved knee function but did not fully resolve symptoms, particularly in the presence of larger defects (>2 cm2) or concurrent meniscal pathology. The study reported that even with isolated debridement, symptoms persisted, and overall pain intensity remained unchanged, highlighting that debridement alone often falls short of providing complete symptom relief. 44 Therefore, clinicians should maintain a high index of suspicion for the need for future implantation in patients with previous surgical history, tailoring preoperative counseling accordingly.
Patellar defects accounted for 65% of implantation cases, reflecting the unique biomechanical challenges of the patellofemoral joint in the presence of symptomatic chondral lesions. Specifically, the patellofemoral joint experiences distinct loading profiles during functional tasks, with peak patellofemoral pressures of 3.3 times body weight when stair climbing, 5.6 times body weight when running, and 7.8 times body weight during deep knee flexion or squatting.15,24 Moreover, chondral defects to the patella differ biomechanically from defects to the femoral condyle or trochlea due to its complex articular geometry characterized by concave medial and lateral facets separated by a convex longitudinal ridge, in addition to the dynamic tracking patterns, which create shear-dominated stress vectors rather than compressive loading.23,42 A cadaveric study reported peak patellofemoral pressures of 6.5 times body weight at 90° of knee flexion, with a 45% increase in lateral facet loading when the Q-angle deviates by 10° from anatomic norms. 17 Anatomic and physiological abnormalities, such as trochlear dysplasia or lateral patellar maltracking, may contribute to elevated rates of patellar cartilage defects requiring MACI implantation. 14
Concomitant procedures performed during the biopsy were associated with a significant reduction in the likelihood of implantation, emphasizing the importance of addressing any joint instability or meniscal pathology, even in the presence of unanticipated chondral lesions that warrant a biopsy. Procedures such as cruciate ligament reconstruction or meniscal repair mitigate aberrant kinematics, thereby reducing shear forces across the chondral surfaces.5,22,32 Namely, meniscal repair has been shown to restore contact area and normalize peak contact pressures, potentially protecting chondral defects from progressive degeneration after adequate debridement.38,46 These findings align with the ICRS recommendations, which have reported that management of chondral defects without treatment of knee instability yields suboptimal results. 29 Assessment and management of injuries to the meniscus, cruciates, or collateral ligaments is crucial, while also addressing any potential joint malalignment that may predispose chondral lesions to further degeneration. While further studies are warranted to determine specific injuries or pathologies predisposing chondral lesions to degenerative progression, addressing all concomitant pathologies at the time of chondral biopsy may help decrease the risk for subsequent chondral implantation. However, it must be recognized that treating these other conditions at the time of the chondral biopsy may necessitate 2 separate and lengthy courses of rehabilitation if the patient ultimately undergoes MACI implantation. This may have practical negative consequences in the ability to return to work and other activities and may be at least partially responsible for the 29% implantation rate that we found.
Patients possessing a chondral defect of ICRS grade 3 or 4 underwent a significantly higher rate of implantation procedures. High-grade lesions, which range from >50% of cartilage depth (grade 3) to full-thickness loss involving subchondral bone (grade 4), are generally less amenable to conservative management compared with lower-grade lesions (ICRS grades 1 and 2). 9 Characterized by higher failure rates and inferior long-term durability after nonoperative treatment, high-grade lesions are associated with a faster progression to osteoarthritis,7,20 with higher levels of pain and functional loss.2,45 Substantial improvements in patient-reported outcomes have been reported, along with lower failure rates, particularly for high-grade lesions after MACI compared with microfracture. 8 In a cohort of 21 patients exclusively with ICRS grade 4 lesions, Kreuz etal 19 reported significant improvements in clinical scores, with radiological assessments showing normal to near-normal values in 74% of assessed parameters after MACI implantation at 12 years. A recent randomized controlled trial showed that patients with full-thickness (ICRS grades 3 and 4) lesions achieved significantly improved Knee injury and Osteoarthritis Outcome Scores after MACI compared with microfracture, at 2 years. 30 In contrast to low-grade partial-thickness lesions, which are less likely to require surgical management,18,20 large, symptomatic, isolated grade 3 and 4 lesions in young, active patients remain a clear surgical indication for MACI. 4
The inverse correlation between mechanical symptoms and progression to phase 2 MACI implantation likely reflects the resolution of unstable chondral lesions through adequate arthroscopic debridement. Mechanical symptoms, such as catching or locking, may arise when large chondral flaps are present, which disrupt normal joint kinematics, impeding smooth joint motion. 31 Arthroscopic debridement of unstable chondral tissue can effectively resolve mechanical symptoms.31,41 This resolution of mechanical symptoms may thereby reduce the need for progression to phase 2 MACI implantation. This finding highlights the importance of precise symptom attribution to guide clinical decision-making by conducting a thorough preoperative evaluation and/or diagnostic arthroscopy to fully ascertain both the subjective symptom pattern and objective anatomic characteristics of the chondral lesions.
This investigation is not without limitations. The retrospective design introduces the potential risk for selection bias. However, no patients undergoing chondral biopsy were indicated for MACI implantation at the time of initial surgery, with implantation only indicated in the presence of persistent, concordant symptoms associated with the known chondral lesions. The single-center nature of the study may limit generalizability, as institutional variations in operative management as well as rehabilitation protocols, such as weightbearing restrictions and extent of nonoperative management for chondral lesions, may influence outcomes. Additionally, variability in operative techniques and nonoperative management strategies among the 5 fellowship-trained surgeons may introduce inconsistencies in both treatment and outcome, although this was not directly evaluated. Chondral defect locations could only be statistically compared per lesion, as many patients had lesions at multiple locations. While none of the defect locations resulted in a significant increase in conversion rate over others, this may represent a confounding variable, given that in the setting of multiple defects, it could not be reliably determined which lesion most contributed to a patient’s decision to undergo implantation. Patient-reported outcome measures were not recorded, and therefore subjective experiences such as pain, function, kinesiophobia, and quality of life were not analyzed. Finally, variability in insurance coverage may have affected the decision to proceed with MACI implantation independent of a patient’s symptoms, physical examination, lesion characteristics, or overall knee function.
Conclusion
The results of this study found that 29% of patients undergoing a MACI biopsy for a chondral lesion progressed to MACI implantation. Patients with patellar defects exhibited a significantly greater likelihood of MACI, with increased conversion rates associated with previous ipsilateral knee surgery and/or more advanced chondral defects. Patients undergoing concomitant procedures at biopsy were significantly less likely to require MACI, with lower conversion rates associated with a history of preoperative mechanical symptoms.
Footnotes
Final version submitted December 1, 2025; accepted December 10, 2025.
One or more of the authors has declared the following potential conflict of interest or source of funding: D.M.K. has received research support from Arthrex and support for education from Elite Orthopaedics and Smith & Nephew. M.V.S. has received consulting fees from Arthrex and support for education from Elite Orthopaedics. M.J.M. has received consulting fees from Arthrex, Breg, and Ostesys; support for education from Elite Orthopaedics; and is the chair of the NFL Research and Innovation Committee. R.H.B. has received consulting fees from Syneos and Anika, and is the chair of the NFL Musculoskeletal Committee. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or responsibility relating thereto.
Ethical approval for this study was obtained from the Washington University Institutional Review Board.
