Clinical and Radiographic Outcomes After Treatment of Patellar Chondral Defects: A Systematic Review

Abstract

Context:

There is currently no evidence-based consensus on how to treat a full-thickness, symptomatic articular cartilage injury of the patella, although numerous treatment options are available.

Objective:

To systematically evaluate the functional outcomes after operative treatment of patellar cartilage lesions. Our secondary purpose was to evaluate radiographic outcomes after treatment.

Data Sources:

PubMed, Cochrane, and Embase

Study Selection:

Studies published between January 1, 1990 and December 31, 2018 that included patient-reported functional outcomes for patients after operative treatment of patellar chondral defects at a minimum 2-year follow-up were included.

Study Design:

Systematic review.

Level of Evidence:

Level 4.

Data Extraction:

MINORS (Methodological Index for Non-Randomized Studies) score, level of evidence, sample size, demographic data, follow-up data, intervention, functional outcome scores, and magnetic resonance imaging (MRI) data were collected.

Results:

The review identified 10 studies and 293 patients receiving cartilage restoration procedures for patellar chondral defects with extractable clinical and radiographic results and data on complications and reoperations. All treatments (autologous chondrocyte implantation [ACI], matrix-induced ACI [MACI], autologous osteochondral transplantation [AOT]) utilized in the management of patellar chondral lesions, with the exception of isolated particulated juvenile articular cartilage, demonstrated statistically significant improvements in functional outcome scores compared with preoperative measurements at a minimum of 2-year follow-up. Postoperative MRIs were obtained in 6 studies and found that regardless of treatment, moderate-to-complete infill of patellar cartilage lesions was seen in the majority of patients. While failure rates were low for the various treatment modalities, rates of reoperation were substantial, with up to 40% to 60% reoperation rate seen after ACI.

Conclusion:

Patients treated with ACI, MACI, and AOT all demonstrated statistically significant improvements in functional outcome scores with radiographic evidence of healing at minimum of 2-year follow-up. Evidence is insufficient to recommend one particular treatment over another.

The patella is the most common location of articular cartilage defects in the knee, seen in 63% of knee arthroscopies.¹¹ The treatment of patellar chondral defects remains challenging, with poorer outcomes and increased failure rates when compared with the treatment of chondral defects of the femoral condyles.^20,24,26 Biomechanically, patellar cartilage lesions have limited intrinsic healing potential due to isolation from systemic regulation, lack of vessels and nerve supply, and sparsely distributed chondrocyte ultrastructure.^8,9 Moreover, the patellofemoral joint sustains significant shear and compression forces, frequently compounded by patellar malalignment, which further complicates the treatment and healing of patellar chondral lesions.^31,32

There is currently no evidence-based consensus on how to treat a full-thickness, symptomatic articular cartilage injury of the patella, although numerous treatment options are available.⁴ Treatment strategies include marrow-stimulation procedures (eg, microfracture, drilling),⁴¹ osteochondral transplantation with autograft or allograft,^23,44 autologous chondrocyte implantation (eg, ACI, matrix-induced ACI [MACI]),⁵ and particulated juvenile articular cartilage allograft tissue.^1,43 However, despite the wide variety of treatment options for chondral lesions about the knee and the broad expanse of published literature on patellofemoral lesions, objective criteria about outcomes after treatment of patellar chondral defects are poorly defined.

The primary purpose of this investigation was to systematically evaluate the literature for functional outcomes after operative treatment of patellar cartilage lesions. Our secondary purpose was to evaluate radiographic outcomes after treatment. We hypothesized that ACI and MACI treatments would be associated with greatest improvement in functional and radiographic outcomes for small- to medium-sized patellar chondral defects.

Methods

Search Strategy

A systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines using a PRISMA checklist.²⁷ All literature published between January 1, 1990 and December 31, 2018 related to cartilage restoration procedures to treat patellar chondral defects were included in our search. Three reviewers, with the assistance of a medical librarian, conducted the search on April 11, 2019 using the following databases: PubMed, Cochrane, and Embase. The search terms utilized during review of the databases are listed in Table 1.

Table 1.

PubMed search

Search No.	Query Input
1	microfracture*[Text Word])
2	autologous chondrocyte implant*[Text Word]) OR aci[Text Word]) OR cartilage implant*[Text Word]) OR matrix induced chondrocyte implant*[Text Word]) OR maci[Text Word])
3	particulated juvenile cartilage allograft*[Text Word])
4	osteochondral autograft transplant*[Text Word]) Word]) OR osteochondral autograft transfer*[Text Word]) OR osteochondral autograft*[Text Word]) OR osteochondral allograft transplant*[Text Word]) OR osteochondral allograft resurfacing[Text Word]) OR osteochondral allograft transfer*[Text Word]) OR osteochondral allograft*[Text Word]) OR oats[Text Word])))
5	1 or 2 or 3 or 4
6	patella[MeSH Terms] OR patella[Text Word] OR patellar[Text Word] OR patellas[Text Word] OR knee cap[Text Word] OR knee caps[Text Word] OR kneecap[Text Word] OR kneecaps[Text Word]
7	(cartilage, articular[MeSH Terms]) OR articular cartilage[Text Word] OR joint cartilage[Text Word] OR chondral[Text Word])
8	fractures, cartilage[MeSH Terms] OR defect*[Text Word] OR injury[Text Word] OR injuries[Text Word] OR injuries[MeSH Subheading] OR damage[Text Word] OR lesion[Text Word] OR lesions[Text Word]
9	5 AND 6 AND 7 AND 8
10	limit 9 to (case reports or editorial or letter or “review”)
11	9 NOT 10
12	limit 11 to animals
13	11 not 12
14	limit 13 to (English language)

The inclusion criteria were studies published in English or with an English translation that included patient-reported functional outcomes for patients after operative treatment of patellar chondral defects. Exclusion criteria were meta-analyses, systematic reviews, review articles, case reports, technique papers, biochemical analyses, cadaveric, in vitro, or animal studies; studies documenting operative intervention of additional chondral defects (eg, femoral, tibial), studies with fewer than 10 patellar defects, and studies with mean follow-up of less than 2 years.

Data Abstraction and Analysis

A total of 97 articles were identified in our systematic search. The search process is shown in the flow diagram (Figure 1). After title and abstract assessment by 2 independent authors, 35 full-text articles were selected for further review. An additional 24 articles were excluded on evaluation of full text. To ensure that all available studies were identified, references within each of the included articles were cross-referenced for inclusion, in case they were overlooked during the initial search. Of the 11 remaining studies that originally met inclusion and exclusion criteria and underwent data extraction, 1 additional study was excluded because it was found to have a mean follow-up less than 2 years, which met our exclusion criteria.⁷ Ultimately, 10 studies were analyzed in this systematic review.^{2,14,16,17,22,29,33,42,43,45}

Figure 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) diagram.

Level of Evidence

The criteria outlined by Wright et al⁴⁷ were used to evaluate the relative merit of the studied included in this review. In addition, the MINORS (Methodological Index for Non-Randomized Studies) criteria outlined by Slim et al³⁹ were used to evaluate each study. MINORS is a well-validated instrument designed to assess the methodological quality of nonrandomized surgical studies, whether comparative or noncomparative. The instrument utilizes 12 questions, which are each given a value of 0 to 2 (0 = not reported, 1 = reported but inadequate, 2 = reported and adequate). The ideal MINORS score for noncomparative studies is 16 and the ideal score for comparative studies is 24. Each article was independently graded by 2 authors (Table 2). Any interreviewer variation, defined as difference in study level or difference in MINORS score >1, was resolved by a third reviewer. Demographic data, defect data, surgical technique, and complication profile were recorded for all studies.

Table 2.

Evaluation of study designs

Study (Year)	Data Collection	Data Abstraction	Level of Evidence	Study Design	Comparative Group	MINORS Score
von Keudell et al (2017)45	Prospective	Retrospective	4	Case series	No	11.5
Gillogly et al (2014)17	Prospective	Retrospective	4	Case series	No	11.5
Teo et al (2013)42	Prospective	Retrospective	4	Case series	No	9.5
Macmull et al (2012)29	Prospective	Retrospective	4	Case series	No	10.5
Henderson and Lavigne (2006)22	Prospective	Retrospective	4	Case series	Yes	11
Ebert et al (2015)14	Prospective	Retrospective	4	Case series	No	14
Astur et al (2014)2	Prospective	Retrospective	4	Case series	No	10.5
Figueroa et al (2011)16	Prospective	Retrospective	4	Case series	No	12
Nho et al (2008)33	Prospective	Retrospective	4	Case series	No	11.5
Tompkins et al (2013)43	Prospective	Retrospective	4	Case series	No	8.5

MINORS, Methodological Index for Non-Randomized Studies.

Functional Outcomes Scores

The main outcome of interest in our review was change in postoperative function, which was measured through patient-reported functional outcomes scores. There was substantial variability in the functional outcomes scores that were reported, with 22 different scales reported in the 10 studies. The most commonly reported functional scores were the International Knee Documentation Committee (IKDC; n = 7) Subjective Knee Form, the Modified Cincinnati Knee Rating System (n = 6), the 36-Item Short Form Survey Instrument (SF-36; n = 5), the visual analogue scale (VAS; n = 4), and the Lysholm Knee Scoring Scale (n = 4). The minimal clinically important difference (MCID) values have been published for the following outcome scores in relation to knee and cartilage surgeries: IKDC (MCID, 10.8),³⁶ Knee Injury and Osteoarthritis Outcome Score (KOOS) Pain (MCID, 11),³⁶ KOOS Symptom (MCID, 3.6),³⁶ KOOS Activities of Daily Living (MCID, 9.2),³⁶ KOOS Sport/Recreation (MCID, 12.5),³⁶ KOOS Quality of Life (MCID, 12.8),³⁶ Knee Society Score (KSS) Knee (MCID, 7.2),²⁸ KSS Function (MCID, 9.7),²⁸ SF-36 (MCID, 10),¹⁵ VAS (MCID, 2.7),²⁵ Lysholm (MCID, 4.2),³⁶ Kujala (MCID, 5),⁴⁶ Tegner Activity Scale (MCID, 1),¹⁰ Modified Cincinnati Knee Score (MCID, 14),⁴⁰ Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) (MCID, 15).¹⁵

Magnetic Resonance Imaging

Our secondary outcome was patellar defect healing on magnetic resonance imaging (MRI). Six studies included postoperative MRI results. The most common methodology of assessing defect healing on MRI was the International Cartilage Repair Society (ICRS) score, which was reported in 3 studies.

Meta-analysis

A true meta-analysis was not possible because of the absence of intrastudy comparison data and variability in study design, surgical technique, and patient-reported outcome measures.

Results

Study Characteristics

The 10 studies meeting inclusion and exclusion criteria were published from 2003 to 2017. All 10 studies were retrospective and level 4 evidence. MINORS scores ranged from 8.5 to 14 (Table 2). All studies had a minimum of 2-year follow-up (range, 24-90.7 months).

Demographics

The 10 studies that met the inclusion criteria consisted of 293 patients receiving cartilage restoration procedures for patellar chondral defects (Table 3). The average study sample size was 29.1 (range, 10-48). The average patient age was 30.33 years (range, 12-61 years). There were 173 (59%) male and 120 (41%) female patients. A total of 12 (4.1%) patients who were lost to follow-up.

Table 3.

Demographic characteristics and study design

Study (Year)	Sample Size, n	Gender (Male/Female)	Age, y, Mean ± SD	Mean Follow-up, mo	Lost to Follow-up, n (%)
von Keudell (2017)45	30	12/18	32 ± 10	88 ± 45	3 (11)
Gillogly et al (2014)17	23 (25 knees)	11/12	31 ± 7	90.7	4 (4)
Teo et al (2013)42	23	19/4	16.8, range 12-21	72	0
Macmull et al (2012)29	48	34/14	34.8, range 17-50	40.3	0
Henderson and Lavigne (2006)22	44	21/23	32.1, range 17-56 (ACI + TTO) 35.1, range 14-55 (ACI alone)	26.2, range 9-52 (ACI + TTO) 28.9, range 11-55 (ACI alone)	0
Ebert et al (2015)14	47 (24 patella)	30/17	37.6, range 20-61	24	2 (4)
Astur et al (2014)2	33	17/16	37.6, range 16-59	Median 30.2, range 24-54	0
Figueroa et al (2011)16	10	10/0	20.2, range 15-38	37.3, range 24-70	0
Nho et al (2008)33	22	12/10	30 ± 12	28.7, range 17.7-57.8	0
Tompkins et al (2013)43	13 (15 knees)	7/6	26.4 ± 9.1	28.8 ± 10.2	3 (19)

ACI, autologous chondrocyte implantation; TTO, tibial tubercle osteotomy.

Surgical Technique

The treatment with the most studies reporting outcomes was ACI (n = 5), followed by autologous osteochondral transplantation (AOT, n = 3), and MACI (n = 2). Other procedures evaluated included patellar realignment tibial tubercle osteotomy (TTO), lateral release, vastus medialis obliquus advancement, trochleoplasty, medial patellofemoral ligament reconstruction, chondral synthesis, microfracture, medial retinaculum repair, elmslie trillat osteotomy, chondroplasty, medial plica resection, lateral retinacular release, and medial plicature. Eight of 10 studies reported a concurrent surgery with the cartilage restoration procedure.

Functional Outcomes

Five studies evaluated functional outcomes after ACI procedures for patellar chondral defects (Table 4). One study evaluated ACI alone while 4 studies included ACI with or without concomitant TTO procedures. The average patellar chondral defect size was not reported for 1 of the 5 studies, but for the other 4 studies ranged from 2.92 to 6.4 cm². The most commonly utilized functional outcome measures were the modified Cincinnati (4) and IKDC scores (3). At latest follow-up, all studies demonstrated statistically significant improvements in functional outcome scores after ACI procedures (Table 4).

Table 4.

Functional outcome scores after treatment of patellar chondral defects

Study (Year)	Patella Defect Sample Size, n	Mean Patella Defect Size, cm2	Treatment	Preoperative Functional Score, Mean ± SD (Range)	Postoperative Functional Score, Mean ± SD (Range)	Change in Functional Score	MRI Data, Mean Months Postoperative
von Keudell et al (2017)45	30	4.7 ± 2.1	ACI ± TTO, soft tissue balancing, trochleoplasty	Modified Cincinnati: 3.1 ± 1.1 KSS-Function: 55.7 ± 12.8 KSS-Pain: 63.9 ± 12.9 WOMAC: 52.2 ± 16.9 SF-36-PCS: 40.0 ± 8.2 SF-36-MCS: 47.0 ± 8.4	Modified Cincinnati: 5.7 ± 1.5 KSS-Function: 73.0 ± 14.7 KSS-Pain: 81.8 ± 12.9 WOMAC: 27.9 ± 23.6 SF-36-PCS: 47.0 ± 10.0 SF-36-MCS: 53.0 ± 8.8	Modified Cincinnati: 2.6 (P < 0.01) KSS-Function: 17.3 (P < 0.01) KSS-Pain: 17.9 (P < 0.01) WOMAC: 14.3 (P < 0.01) SF-36-PCS: 7.0 (P = 0.01) SF-36-MCS: 6.0 (P = 0.02)	Yes (24)
Gillogly et al (2014)17	25	6.4	ACI + TTO ± Trochleoplasty	Modified Cincinnati: 3.0 IKDC: 43 Lysholm Score: 40 SF-12 (Mental): 48 SF-12 (Physical): 41	Modified Cincinnati: 7.0 IKDC: 76 Lysholm Score: 79 SF-12 (Mental): 61 SF-12 (Physical): 48	Modified Cincinnati: 4.0 (P < 0.0001) IKDC 33.2 (P < 0.0001) Lysholm Score: 39.1 (P < 0.0001) SF-12 (Mental): 12.6 (P = 0.0001) SF-12 (Physical): 6.4 (P = 0.002)	No
Teo et al (2013)42	23	Not reported	ACI ± TTO (20) BMSC ± TTO (3)	IKDC: 45 (range, 5.8-63) Lysholm Gilquist: 50 (range, 11-79) Tegner Lysholm: 2.5 (range, 0-5)	IKDC: 75 (range, 40.2-96.6) Lysholm Gilquist: 70 (range, 48-100) Tenger Lysholm: 4 (range, 2-7)	IKDC: 30 (P < 0.001) Lysholm Gilquist: 20 (P < 0.001) Tegner Lysholm: 1.5 (P < 0.001)	No
Macmull et al (2012)29	25	4.73	ACI	VAS: 6.32 (range, 2-10) Stanmore/Bentley: 3.04 (range, 1-4) Modified Cincinnati: 42.12 (range, 18-60)	VAS: 5 (range, 1-9) Stanmore/Bentley: 2.44 (range, 0-4) Modified Cincinnati: 48.76 (range, 11-83)	VAS: −1.32 (P = 0.02) Stanmore/Bentley: −0.6 (P = 0.01) Modified Cincinnati: 6.64 (P = 0.02)	No
Henderson and Lavigne (2006)22	22	2.92	ACI + proximal and distal extensor realignment	Raw values not presented	Raw values not presented	Modified Cincinnati: 4.46 SF-36 PCS: 17.7 (P < 0.001) IKDC: 36.2 (P < 0.001)	No
Henderson and Lavigne (2006)22	22	3.22	ACI	Raw values not presented	Raw values not presented	Modified Cincinnati: 1.73 SF-36 PCS: 12.1 (P < 0.05) IKDC: 22.3 (P < 0.05)	No
Ebert et al (2015)14	24	MACI + TTO: 3.4 (1.0-5.0) MACI alone: 3.1 (2.3-6.0)	MACI ± TTO	KOOS-Pain: 61.4 ± 15.6 KOOS-Symptom: 64.7 ± 17.2 KOOS-ADLs: 69.0 ± 16.1 KOOS-Sport and Recreation: 24.6 ± 21.0 KOOS-Quality of Life: 22.9 ± 16.0 36-Item Short Form Health Surgery SF-36-PCS: 34.9 ± 9.7 SF-36-MCS: 51.3 ± 8.8 VAS: 5.4 ± 1.4	KOOS-Pain: 83.3 ± 11.4 KOOS-Symptom: 86.4 ± 9.8 KOOS-ADLs: 87.5 ± 11.0 KOOS-Sport and Recreation: 50.1 ± 29.4 KOOS-Quality of Life: 53.3 ± 23.0 SF-36-PCS: 47.1 ± 10.2 SF-36-MCS: 55.7 ± 6.1 VAS: 1.8 ± 1.1	KOOS-Pain: 21.9 (P < 0.001) KOOS-Symptom: 21.7 (P < 0.001) KOOS-ADLs: 18.6 (P < 0.001) KOOS-Sport and Recreation: 25.5 (P < 0.001) KOOS-Quality of Life: 30.4 (P < 0.001) SF-36-PCS: 12.2 (P < 0.01) SF-36-MCS: 4.4 (P = 0.002) VAS: −3.6 (P < 0.001)	Yes (24)
Macmull et al (2012)29	23	4.76	MACI	VAS: 6.52 (range, 2.5-10) Stanmore/Bentley: 2.78 (range, 1-4) Modified Cincinnati: 48.39 (range, 22-75)	VAS: 3.96 (range, 0-10) Stanmore/Bentley: 2.09 (range, 0-4) Modified Cincinnati: 61.39 (range, 8-100)	VAS: −2.56 (P < 0.001) Stanmore/Bentley: −0.69 (P < 0.001) Modified Cincinnati: 19.27 (P < 0.01)	No
Astur et al (2014)2	33	Not reported; all defects between 1 and 2.5 cm in diameter	AOT	Lysholm: 57.27 ± 19.97 Fulkerson: 54.24 ± 18.89 Kujala: 54.76 ± 17.61 SF-36-PCS: 45.91 ± 13.31 SF-36-MCS: 65.58 ± 17.03	Lysholm: 80.76 ± 12.26 Fulkerson: 80.42 ± 10.20 Kujala: 75.18 ± 12.47 SF-36-PCS: 63.64 ± 29.11 SF-36-MCS: 74.06 ± 21.24	Lysholm: 23.49 (P < 0.001) Fulkerson: 26.18 (P < 0.001) Kujala: 20.42 (P < 0.001) SF-36-PCS: 17.73 (P = 0.01) SF-36-MCS: 8.48 (P = 0.05)	Yes (12)
Figueroa et al (2011)16	10	1.2	AOT ± MPFL, TTO, soft tissue balancing	Lysholm: 73.8 ± 8.36 (range, 66-86) IKDC: not reported	Lysholm: 95 ± 4.47 (range, 90-100) IKDC: 93.6 ± 1.74 (range, 92-96)	Lysholm: 21.2 (P <0.05) IKDC: n/a	Yes (8)
Nho et al (2008)33	22	1.65 ± 1.3	AOT ± TTO, lateral release, distal realignment, proximal realignment	IKDC 47.2 ± 14.0 (range, 20.7-71.3) ADL 60.1 ± 16.9 (range, 23.8- 93.8 SF-36 64.0 ± 14.8 (range, 39.1-85.6)	IKDC 74.4 ± 12.3 (range, 51.7-87.4) ADL 84.7 ± 8.3 (range, 65.0-93.7) SF-36 79.4 ± 15.4 (range, 43.8-92.0)	IKDC: 27.2 (P = 0.03) ADL: 24.6 (P = 0.02) SF-36: 15.4 (P = 0.06)	Yes (17.3)
Tompkins et al (2013)43	15	2.4 ± 1.2	Particulated juvenile articular cartilage allograft ± AMZ, MPFL	Tegner Score: 7 (range, 3-9; consistent with competitive sports)	Tegner: 5 (range, 3-9; consistent with return to recreational sports or heavy labor) KOOS-Pain: 84.2 ± 12.4 KOOS-Symptoms/Stiffness: 85.0 ± 12.3 KOOS-ADL: 88.9 ± 12.9 KOOS-Sports and Recreation: 62.0 ± 25.1 KOOS-QOL: 60.8 ± 28.6 IKDC: 73.3 ± 17.6 Kujala: 79 (range, 55-99) VAS: 1.9 ± 1.4 Subjective return to presymptomatic activity: 5/13 (38.5%) patients	Tegner Score: −2	Yes (28.8)

ACI, autologous chondrocyte implantation; ADL, activities of daily living; AMZ, anteromedialization of tibial tubercle; AOT, autologous osteochondral transplantation; BMSC, bone-marrow stem cell; IKDC, International Knee Documentation Committee Subjective Knee Form; KOOS, Knee Injury and Osteoarthritis Outcome Score; MACI, matrix-induced autologous chondrocyte implantation; MPFL, medial patellofemoral ligament; n/a, not applicable; SF-12, 12-Item Short Form Survey Instrument; SF-36, 36-Item Short Form Survey Instrument; TTO, tibial tubercle osteotomy; VAS, visual analogue scale; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index.

Two studies examined MACI procedures with or without TTO for cartilage defects of the patella. The mean patellar chondral defect size was 3.75 cm². At latest follow-up, statistically significant improvements in all functional outcome scores (KOOS, SF-36, VAS, Stanmore/Bentley, and modified Cincinnati) were seen in both studies.

Our review captured 3 studies investigating AOT with or without TTO for the treatment of patellar chondral defects. The mean patellar chondral defect size was 1.43 cm² for 2 of the included studies and was not reported for 1 of the studies (instead a range from 1 to 2.5 cm in diameter). All 3 studies reported statistically significant improvements in functional outcome scores at latest follow-up.

Only 1 study evaluating isolated particulated juvenile articular cartilage allografts in patellar cartilage defects met inclusion criteria. The average patellar chondral defect size in the study by Tompkins et al⁴³ was 2.4 cm². While numerous postoperative functional scores are presented in Table 4, only the Tegner activity score was available for comparison before and after surgical intervention, and interestingly, demonstrated a slight decrease from a score of 7 preoperatively (corresponding to competitive level sports) to a score of 5 (corresponding to work-heavy labor level of activity) at a mean of 28.8 months postoperatively.

MRI Findings

Six of the 10 studies included postoperative MRI data to assess patellar cartilage repair. The mean time to postoperative MRI was 19 months across the 6 studies (range, 8-28.8 months). The most commonly reported parameter was the percentage of cartilage infill (5/6 studies). After ACI with or without TTO, von Keudell et al⁴⁵ reported complete infill of patellar cartilage defects in 75% of their patients at 24-month follow-up. Moderate infill was seen in another 17% of their patients.⁴⁵

Similarly, Ebert et al¹⁴ performed MRIs in patients 24 months after MACI procedures with or without TTO. Complete graft infill was seen in 40.4% of their patients, with 78.7% demonstrating infill >50%.¹⁴ In addition, an MRI composite score, based on the magnetic resonance observation of cartilage repair tissue scoring system,³⁰ demonstrated statistically significant improvement over time.

Three studies utilized postoperative MRI for evaluation of cartilage repair after AOT. Cartilage repair infill ranged from 67% to 100% at 17.3 months postoperatively³³ to 100% osseous integration at 12 months.² The ICRS grading system⁶ was utilized in 2 of the AOT studies. Figueroa et al¹⁶ found grade 1 (nearly normal) cartilage in 100% (all 10 patients) at 8 months postoperatively. Nho et al³³ reported grade 0 (normal) cartilage in 7% (1/14) and grade 1 cartilage in 50% (7/14) of their patients at 17.3 months postoperatively.³³

After implantation of isolated particulated juvenile articular cartilage allograft, Tompkins et al⁴³ demonstrated mean cartilage infill to be 89% at a mean of 28.8 months postoperatively.⁴³ In addition, 73% (11/15) of patients who underwent treatment with particulated allograft had normal or near normal ICRS scores.

Failures/Reoperation

Treatment failures were reported for 8 of the 10 studies (Table 5). Failure rates after ACI procedures were reported by Gillogly et al¹⁷ to be 4% (1/25) and by von Keudell et al⁴⁵ to be 10% (3/30) with the failures being subsequent patellofemoral (2) and bicompartmental arthroplasties (1). Henderson and Lavigne²² reported no failures in their study of 44 patients treated with ACI.²² Rates of reoperation were high after ACI procedures, ranging from 40% to 60% in the 3 studies that included this data.^17,22,45 The most common reasons for reoperation after ACI were for graft hypertrophy and removal of hardware (Table 5).

Table 5.

Treatment failures and reoperations after treatment of patellar chondral defects

Study (Year)	Treatment	Failure Rate	Reason for Failure/Comments	Reoperation Rate	Reason for Reoperation/Comments
von Keudell et al (2017)45	ACI ± TTO, soft tissue balancing, trochleoplasty	10% (3/30)	PF joint replacement: 2 bicompartmental arthroplasty:1	60% (18/30)	Graft hypertrophy: 7 Chondroplasty: 5 Arthrofibrosis: 4 Removal of hardware: 2
Gillogly et al (2014)17	ACI + AMZ ± Trochleoplasty	4% (1/25)	PF arthroplasty: 1	40% (10/25); 15 subsequent procedures	Graft hypertrophy and hardware removal: 5 Graft hypertrophy: 3 Graft hypertrophy, lateral release, excision of neuroma: 1 Loose body removal and chondroplasty: 1
Teo et al42 (2013)	ACI ± TTO (20) BMSC ± TTO (3)	Not reported	N/A	Not reported	N/A
Macmull et al (2012)29	ACI (25) MACI (23)	Not reported	N/A	Not reported	N/A
Henderson and Lavigne (2006)22	ACI ± proximal and distal extensor realignment	0	N/A	52% (23/44)	Graft hypertrophy or extrusion: 23 Removal of hardware: 2 Meniscectomy: 1 New chondral defect: 1 Fat pad adhesions: 1
Ebert et al (2015)14	MACI ± TTO	4.3% (2/47)	Graft delamination	0	N/A
Astur et al (2014)2	AOT	0	N/A	9% (3/33)	Arthrofibrosis: 3
Figueroa et al (2011)16	AOT ± MPFL, TTO, soft tissue balancing	0	N/A	0	N/A
Nho et al (2008)33	AOT ± TTO, lateral release, distal realignment, proximal realignment	4.5% (1/22)	Very large defect (500 mm2)	18% (4/22)	Hardware removal: 3 Arthroscopic debridement and repeat lateral release: 1
Tompkins et al (2013)43	Particulated juvenile articular cartilage ± AMZ, MPFL	0	N/A	33% (5/15)	Arthrofibrosis: 2 Arthroscopic debridement of graft hypertrophy: 2 Arthroscopic debridement of partially filled defect: 1

ACI, autologous chondrocyte implantation; AMZ, anteromedialization of tibial tubercle; AOT, autologous osteochondral transplantation; BMSC, bone-marrow stem cell; MACI, matrix-induced autologous chondrocyte implantation; MPFL, medial patellofemoral ligament; N/A, not applicable; PF, patellofemoral; TTO, tibial tubercle osteotomy.

Only 1 study, by Ebert et al,¹⁴ reported on the failures and complications after MACI. They reported a 4.3% (2/47) failure rate due to graft delamination and/or exposed subchondral bone on MRI, although there was no mention of reoperations within their study period.

After AOT for the treatment of patellar chondral defects, only 1 treatment failure, 1.5% (1/66) was noted out of 3 studies.^3,16,33 This treatment failure was noted to be in a patient with a very large >5 cm² chondral defect.³³ There was a 10.8% (7/65) reoperation rate, the most common reasons being for arthrofibrosis (3) and hardware removal (3).

Discussion

Articular cartilage defects remain a challenging problem, particularly with regard to the patella, with poorer outcomes and increased failure rates when compared with the treatment of chondral defects of the femoral condyles.^20,24,26

Functional Outcomes

The findings of this systematic review, evaluating all available studies with functional outcomes scores specific to patellar cartilage defects and a minimum of 2-year follow-up, highlight the limited number of studies available to direct clinical decision making related to patellar lesions. Overall, all treatments (ACI, MACI, AOT) utilized in the management of patellar chondral lesions, with the exception of isolated particulated juvenile articular cartilage, demonstrated statistically significant improvements in functional outcome scores compared with preoperative measurements at a minimum of 2-year follow-up. It is important to note that while most improvements in patient-reported outcomes scores did achieve statistical significance, score improvements did not always reach the MCIDs that have been published for the different measurement tools. Given the high degree of variability in patient-reported outcome measures, it is difficult to compare patient outcomes across treatment groups.

Patellofemoral Alignment

Patellofemoral alignment is thought to play a critical role in the initiation and progression of patellofemoral chondral pathology and its response to operative interventions.⁵ The decision of when to perform realignment of the extensor mechanism is complex and based on multiple factors including the location of patellofemoral cartilage loss, positioning of the patella relative to the femoral trochlea, and the relationship of the tibial tuberosity relative to the trochlear groove.³⁷ Unfortunately, few studies have directly addressed the effectiveness of combined distal realignment associated with cartilage repair in isolated patellar cartilage lesions. Henderson and Lavigne²² directly compared ACI in the presence or absence of TTO and reported 86% good to excellent outcomes in the group treated with ACI with realignment versus 55% treated with ACI alone. This difference was postulated to be due to the unloading effect of the osteotomy as well as incorrect preoperative evaluation of the extent of malalignment in the patients who underwent ACI alone. Interestingly, they concluded that patellofemoral joint unloading with a realignment procedure is desirable to maximize the clinical outcome of ACI, even when no tracking anomalies are identified clinically.

Radiographic Findings/MRI

Although arthroscopy is often considered the gold standard in diagnosing cartilage defects by direct visualization, MRI has become an invaluable tool in detecting, staging, and monitoring patellar chondral lesions.^12,13 The 6 included studies that performed postoperative MRI demonstrate that regardless of treatment, the vast majority of patients demonstrate moderate to complete infill of patellar cartilage lesions with “near normal” appearance at latest follow-up. However, aside from graft infill and gross appearance, Nho et al³³ performed a more detailed analysis of their MRIs in patients after AOT and found that, despite appropriate infill, there was a mismatch between the subchondral plate of the plug and the surrounding native patella with some degree of fissuring in 100% of cases. This is notable as the chondral portion of the osteochondral plug did not appear to show evidence of peripheral healing over time. Although the clinical relevance of these findings remains to be seen, this observation highlights the technical difficulties of transplanting or treating patellar chondral lesions to recreate the native surface geometry, cartilage thickness, and subchondral bone modulus to create a mechanical and biologic environment most conducive to cartilage incorporation.

Failures/Reoperations

While overall failures for the various treatments of patella chondral defects were relatively low, rates of reoperation in the included study were substantial. The highest rates of reoperation were seen after ACI procedures, with the most common reasons for reoperation being graft hypertrophy and removal of hardware. Graft hypertrophy is a well-documented complication after ACI, reported to occur in up to 30% to 40% of patients.²¹ While early reports associated graft hypertrophy with the use of autologous periosteum as described as “first-generation ACI,” graft hypertrophy has also been described in patients after second- or even third-generation ACI.^18,19,34,38 The exact etiology of graft hypertrophy remains elusive, although multiple factors including patient body mass index, age, defect size, and defect location have been proposed. More recent work has suggested that graft hypertrophy may be a temporary adaptation process to the static compressive forces in the knee and that full maturation of the ACI grafts may take up to 2-4 years.³⁵

Limitations

The findings of this study must be interpreted within the limitations of the current literature. The small sample size of included studies analyzed was related to the strict inclusion and exclusion criteria used. All the included studies consisted of small retrospective case series. Functional outcomes scores used to assess outcomes after the various treatments for patellar chondral lesions were extremely variable with at least 22 different outcome measures reported in our review. This variability made it difficult to compare outcomes between studies and across treatment modalities, and because of the heterogeneity, meta-analysis could not be performed.

Conclusion

Patellar cartilage defects remain a challenging orthopaedic problem. The results of our systematic review demonstrate significant improvements in patient-reported functional outcomes scores with use of ACI, MACI, and AOT for the treatment of patellar chondral defects. There was insufficient patient outcome data for marrow stimulation procedures and particulated juvenile articular cartilage allograft tissue for patellar cartilage lesions. Rates of reoperation, particularly after ACI procedures, were high and were most commonly performed for graft hypertrophy and hardware removal. Further high-quality prospective studies are needed to better compare functional outcomes for patellar cartilage lesions across treatment groups, particularly with regards to newer cartilage repair techniques, and with emphasis on cost-effectiveness.