Advertisement

Revision anterior cruciate ligament reconstruction does not prevent progression in one out of five patients of osteoarthritis: a meta-analysis of prevalence and progression of osteoarthritis

      Purpose

      To summarise all the up-to-date evidence related to knee osteoarthritis (OA) after revision anterior cruciate ligament (ACL), and to determine the overall average postoperative OA, its progression and its prevalence in each knee compartment after a minimum 4 years, independent from patients' characteristics and surgical techniques.

      Methods

      A systematic review was performed in PubMed and EMBASE. All the studies that reported knee OA at preoperative status, at final follow-up or both were considered for the meta-analysis. A random-effect meta-analysis was used to obtain the incidence and CIs of postoperative OA, OA progression from preoperative status to final follow-up, and OA prevalence in each compartment. Cut-off values for all OA-grading scales were used to dichotomise the outcomes in a normal knee or knee with OA signs.

      Results

      Overall, 19 studies reported the radiographic outcomes after revision ACL in 776 patients, at a mean pooled follow-up of 5.8 years. Age at surgery was 30.2 years. The overall postoperative OA incidence was 61% (CI 50% to 70%). Progression from a normal knee to OA was reported in 20% (CI 10% to 32%) of the participants. Considering single compartments, OA incidence was 46% (CI 41% to 52%) in the medial compartment, 23% (CI 13% to 36%) in the lateral compartment and 27% (CI 18% to 38%) in the patellofemoral compartment.

      Conclusions

      Knee OA was present in almost 60% of the patients at a mean of 6.2 years follow-up after ACL revision, with an almost double-fold incidence in the medial compartment compared to both lateral compartment and patellofemoral compartment. Progression from normal status to OA status was reported in about 20% of cases. These findings highlight the OA as a concrete problem after revision ACL surgery, which could affect clinical results, sport participation and daily life activities.
      What is already known
      • Revision anterior cruciate ligament (ACL) reconstruction presents a higher incidence of meniscal lesions, cartilage damage and, potentially, evolution to osteoarthritis.
      • The incidence of knee osteoarthritis after revision ACL reconstruction, its progression and the distribution within each compartment is not clear.
      What are the new findings
      • Knee osteoarthritis is present in almost 60% of the patients 6.2 years after revision anterior cruciate ligament (ACL) reconstruction.
      • In one of the five patients who underwent revision ACL reconstruction, osteoarthritis progression is noted.
      • Knee osteoarthritis after revision ACL reconstruction affects the medial, lateral and patellofemoral compartments in 46%, 23% and 27% of the patients, respectively.

      Introduction

      The prophylactic effect of anterior cruciate ligament (ACL) reconstruction in preventing knee osteoarthritis (OA) represents a historically debated issue. A recent meta-analysis defined the risk of developing knee OA after ACL injury as fivefold compared with the normal knee, with slight differences between ACL reconstructed and non-operatively treated patients.
      • Ajuied A
      • Wong F
      • Smith C
      • et al.
      Anterior cruciate ligament injury and radiologic progression of knee osteoarthritis: a systematic review and meta-analysis.
      However, non-surgical treatment is generally indicated for less-active patients who are asked to avoid high-impact activities, while ACL reconstruction is advocated for young athletes who are, in contrast, subjected to high loads and stress on the reconstructed knee.
      ACL reconstruction failure, especially in active and athletic patients, definitely represents a major trauma for the knee joint that could, if untreated, determine a status of chronic instability which can put the menisci and cartilage at risk of further damage. On the other hand, ACL revision surgery also represents an invasive procedure that could further damage and weaken the knee joint due to the necessity of graft harvest, multiple tunnel drilling and hardware removal. In this regard, in the current literature, the radiographic outcome after revision ACL reconstruction has been investigated in several studies with different follow-up, grafts, surgical techniques and concomitant procedures reporting a incidence of postoperative OA ranging from 26% to 100%;
      • Fules PJ
      • Madhav RT
      • Goddard RK
      • et al.
      Revision anterior cruciate ligament reconstruction using autografts with a polyester fixation device.
      ,
      • Gifstad T
      • Drogset JO
      • Viset A
      • et al.
      Inferior results after revision ACL reconstructions: a comparison with primary ACL reconstructions.
      ,
      • Dejour D
      • Saffarini M
      • Demey G
      • et al.
      Tibial slope correction combined with second revision ACL produces good knee stability and prevents graft rupture.
      ,
      • Ventura A
      • Legnani C
      • Terzaghi C
      • et al.
      Revision surgery after failed ACL reconstruction with artificial ligaments: clinical, histologic and radiographic evaluation.
      however, these rates generally derive from heterogeneous case series that involve a small number of patients, various grading systems and different evaluation modalities. The aim of the present meta-analysis was to sum up all the available evidence of knee OA incidence and progression after revision ACL surgery, and to evaluate the changes in the different knee compartments as well.

      Material and methods

      Search strategy

      A systematic review was performed up to 1 October 2015 in the main electronic databases PubMed and EMBASE, using the key words ‘anterior cruciate ligament revision’, ‘ACL revision’, ‘anterior cruciate ligament reoperation’ and ‘ACL reoperation’. All the clinical studies that evaluated the outcomes after revision ACL reconstruction were extracted from the search results and full texts were obtained. Experimental in vitro studies and systematic or narrative reviews were excluded. The reference sections of the main relevant papers were screened in order to find relevant papers eventually missed during the main search. The full text of the selected papers was screened to rule out all clinical studies that did not meet the following inclusion criteria:
      • Prospective or retrospective study design
      • Revision ACL reconstruction isolated or combined with osteotomy, meniscal, cartilage or other ligament surgery
      • Radiographic evaluation of postoperative OA independent from the grading scale
      • Minimum follow-up of 4 years
      Further papers were excluded on the basis of the following exclusion criteria:
      • Incomplete radiographic evaluation (eg, only incidence of OA progression or only difference between preoperative and postoperative status)
      • Review papers
      • Studies not reported in the English language.

      Data extraction

      All relevant information regarding the patient's demographic details was obtained. Follow-up length, graft used for revision reconstruction, concomitant procedures, meniscal and cartilage status were collected as well when available.
      The results from the radiological assessments were extracted from the included studies and divided in a dichotomous manner (no OA signs vs OA) based on the cut-off values for the different radiological classification systems.
      • Øiestad BE
      • Engebretsen L
      • Storheim K
      • et al.
      Knee osteoarthritis after anterior cruciate ligament injury: a systematic review.
      OA was considered in case of Fairbanks grade ≥1, International Knee Documentation Committee (IKDC) 2000 grade ≥B, Kellgren and Lawrence ≥2, Ahlbäck grade ≥1, Jaeger and Wirth grade ≥1, joint space narrowing ≥50%. The overall postoperative incidence of knee OA was based on the previous cut-off values. Studies with OA grading reported for each single compartment (medial, lateral and patellofemoral) were excluded from this main analysis due to the impossibility of determining the correct number of osteoarthritic knees without underestimating or overestimating the results. However, these studies were used to perform a separate analysis to evaluate single compartment postoperative OA incidence.
      A separate analysis was performed, extracting the studies that reported preoperative and postoperative radiographic outcomes. In this case, the incidence of preoperative and postoperative OA was calculated, in addition to the incidence of OA progression. When possible, the rates of OA severity according to each different grading scale were estimated.

      Statistical analysis

      Statistical analysis was performed with MedCalc software (MedCalc software, Acacialaan 22, Ostend, Belgium). Data of the overall postoperative knee OA based on cut-off values and the incidence of OA prevalence for each compartment were calculated through a random-effect model meta-analysis. The prevalence of coupled preoperative and postoperative knee OA, and the incidence of OA progression were calculated in a similar manner. The random-effect model was used to reduce bias from the potential systematic error of the included studies. The inverse variance method was used for the weighting of each study. Homogeneity across the studies was assessed and represented by I2, with p<0.05 being statistically significant.
      • Hedges LV
      • Vevea JL
      Fixed-and random-effects models in meta-analysis.
      • Higgins JP
      • Thompson SG
      • Deeks JJ
      • et al.
      Measuring inconsistency in metaanalyses.
      The pooled patients' age, follow-up length, meniscus or cartilage lesion and the percentage of OA severity according to each grading scale were calculated by a classical pooled analysis. Pooled SD of age and follow-up were not calculated due to the incompleteness of the data of most of the studies.

      Results

      The abstracts of 645 studies from systematic searches were reviewed. According to the inclusion and exclusion criteria, 63 papers reporting the outcomes of revision ACL reconstruction were considered eligible for an initial screening. After further screening, 27 papers were found to report the radiographic outcomes of revision ACL reconstruction. Eight papers were excluded because they did not report an adequate evaluation of knee OA or sufficient follow-up. Therefore, 19 studies
      • Fules PJ
      • Madhav RT
      • Goddard RK
      • et al.
      Revision anterior cruciate ligament reconstruction using autografts with a polyester fixation device.
      ,
      • Gifstad T
      • Drogset JO
      • Viset A
      • et al.
      Inferior results after revision ACL reconstructions: a comparison with primary ACL reconstructions.
      ,
      • Dejour D
      • Saffarini M
      • Demey G
      • et al.
      Tibial slope correction combined with second revision ACL produces good knee stability and prevents graft rupture.
      ,
      • Ventura A
      • Legnani C
      • Terzaghi C
      • et al.
      Revision surgery after failed ACL reconstruction with artificial ligaments: clinical, histologic and radiographic evaluation.
      • Ahn JH
      • Lee YS
      • Chang MJ
      • et al.
      Analysis of revision anterior cruciate ligament reconstruction according to the combined injury, degenerative change, and MRI findings.
      ,
      • Ahn JH
      • Lee YS
      • Ha HC
      Comparison of revision surgery with primary anterior cruciate ligament reconstruction and outcome of revision surgery between different graft materials.
      ,
      • Battaglia II, MJ
      • Cordasco FA
      • Hannafin JA
      • et al.
      Results of revision anterior cruciate ligament surgery.
      ,
      • Diamantopoulos AP
      • Lorbach O
      • Paessler HH
      Anterior cruciate ligament revision reconstruction: results in 107 patients.
      ,
      • Ferretti A
      • Conteduca F
      • Monaco E
      • et al.
      Revision anterior cruciate ligament reconstruction with doubled semitendinosus and gracilis tendons and lateral extra-articular reconstruction.
      ,
      • Fox JA
      • Pierce M
      • Bojchuk J
      • et al.
      Revision anterior cruciate ligament reconstruction with nonirradiated fresh-frozen patellar tendon allograft.
      ,
      • Franceschi F
      • Papalia R
      • Del Buono A
      • et al.
      Two stage procedure in anterior cruciate ligament revision surgery: a five-year follow-up prospective study.
      ,
      • Garofalo R
      • Djahangiri A
      • Siegrist O
      Revision anterior cruciate ligament reconstruction with quadriceps tendon-patellar bone autograft.
      ,
      • Griffith TB
      • Allen BJ
      • Levy BA
      • et al.
      Outcomes of repeat revision anterior cruciate ligament reconstruction.
      ,
      • Grossman MG
      • ElAttrache NS
      • Shields CL
      • et al.
      Revision anterior cruciate ligament reconstruction: three- to nine-year follow-up.
      ,
      • Kievit AJ
      • Jonkers FJ
      • Barentsz JH
      • et al.
      A cross-sectional study comparing the rates of osteoarthritis, laxity, and quality of life in primary and revision anterior cruciate ligament reconstructions.
      ,
      • Mayr HO
      • Willkomm D
      • Stoehr A
      • et al.
      Revision of anterior cruciate ligament reconstruction with patellar tendon allograft and autograft: 2- and 5-year results.
      ,
      • Salmon LJ
      • Pinczewski LA
      • Russell VJ
      • et al.
      Revision anterior cruciate ligament reconstruction with hamstring tendon autograft: 5- to 9-year follow-up.
      ,
      • Thomas NP
      • Kankate R
      • Wandless F
      • et al.
      Revision anterior cruciate ligament reconstruction using a 2-stage technique with bone grafting of the tibial tunnel.
      ,
      • Wirth CJ
      • Kohn D
      Revision anterior cruciate ligament surgery: experience from Germany.
      were included in the systematic review. Details of the exclusion process are depicted in the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow chart for systematic review and meta-analysis (figure 1).
      Figure thumbnail gr1
      Figure 1PRISMA flow chart of the studies included in the systematic review and meta-analyses (ACL, anterior cruciate ligament; OA, osteoarthritis; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses).
      Among the selected studies, 15 studies
      • Fules PJ
      • Madhav RT
      • Goddard RK
      • et al.
      Revision anterior cruciate ligament reconstruction using autografts with a polyester fixation device.
      ,
      • Gifstad T
      • Drogset JO
      • Viset A
      • et al.
      Inferior results after revision ACL reconstructions: a comparison with primary ACL reconstructions.
      ,
      • Dejour D
      • Saffarini M
      • Demey G
      • et al.
      Tibial slope correction combined with second revision ACL produces good knee stability and prevents graft rupture.
      ,
      • Ventura A
      • Legnani C
      • Terzaghi C
      • et al.
      Revision surgery after failed ACL reconstruction with artificial ligaments: clinical, histologic and radiographic evaluation.
      • Ahn JH
      • Lee YS
      • Chang MJ
      • et al.
      Analysis of revision anterior cruciate ligament reconstruction according to the combined injury, degenerative change, and MRI findings.
      ,
      • Ahn JH
      • Lee YS
      • Ha HC
      Comparison of revision surgery with primary anterior cruciate ligament reconstruction and outcome of revision surgery between different graft materials.
      ,
      • Battaglia II, MJ
      • Cordasco FA
      • Hannafin JA
      • et al.
      Results of revision anterior cruciate ligament surgery.
      ,
      • Diamantopoulos AP
      • Lorbach O
      • Paessler HH
      Anterior cruciate ligament revision reconstruction: results in 107 patients.
      ,
      • Ferretti A
      • Conteduca F
      • Monaco E
      • et al.
      Revision anterior cruciate ligament reconstruction with doubled semitendinosus and gracilis tendons and lateral extra-articular reconstruction.
      • Franceschi F
      • Papalia R
      • Del Buono A
      • et al.
      Two stage procedure in anterior cruciate ligament revision surgery: a five-year follow-up prospective study.
      ,
      • Garofalo R
      • Djahangiri A
      • Siegrist O
      Revision anterior cruciate ligament reconstruction with quadriceps tendon-patellar bone autograft.
      ,
      • Griffith TB
      • Allen BJ
      • Levy BA
      • et al.
      Outcomes of repeat revision anterior cruciate ligament reconstruction.
      • Kievit AJ
      • Jonkers FJ
      • Barentsz JH
      • et al.
      A cross-sectional study comparing the rates of osteoarthritis, laxity, and quality of life in primary and revision anterior cruciate ligament reconstructions.
      • Salmon LJ
      • Pinczewski LA
      • Russell VJ
      • et al.
      Revision anterior cruciate ligament reconstruction with hamstring tendon autograft: 5- to 9-year follow-up.
      • Wirth CJ
      • Kohn D
      Revision anterior cruciate ligament surgery: experience from Germany.
      reported the postoperative radiographic results, 9 studies
      • Dejour D
      • Saffarini M
      • Demey G
      • et al.
      Tibial slope correction combined with second revision ACL produces good knee stability and prevents graft rupture.
      • Ventura A
      • Legnani C
      • Terzaghi C
      • et al.
      Revision surgery after failed ACL reconstruction with artificial ligaments: clinical, histologic and radiographic evaluation.
      • Ahn JH
      • Lee YS
      • Chang MJ
      • et al.
      Analysis of revision anterior cruciate ligament reconstruction according to the combined injury, degenerative change, and MRI findings.
      • Ahn JH
      • Lee YS
      • Ha HC
      Comparison of revision surgery with primary anterior cruciate ligament reconstruction and outcome of revision surgery between different graft materials.
      • Diamantopoulos AP
      • Lorbach O
      • Paessler HH
      Anterior cruciate ligament revision reconstruction: results in 107 patients.
      • Franceschi F
      • Papalia R
      • Del Buono A
      • et al.
      Two stage procedure in anterior cruciate ligament revision surgery: a five-year follow-up prospective study.
      • Griffith TB
      • Allen BJ
      • Levy BA
      • et al.
      Outcomes of repeat revision anterior cruciate ligament reconstruction.
      • Salmon LJ
      • Pinczewski LA
      • Russell VJ
      • et al.
      Revision anterior cruciate ligament reconstruction with hamstring tendon autograft: 5- to 9-year follow-up.
      • Wirth CJ
      • Kohn D
      Revision anterior cruciate ligament surgery: experience from Germany.
      reported preoperative and postoperative results, and 8 studies
      • Battaglia II, MJ
      • Cordasco FA
      • Hannafin JA
      • et al.
      Results of revision anterior cruciate ligament surgery.
      • Ferretti A
      • Conteduca F
      • Monaco E
      • et al.
      Revision anterior cruciate ligament reconstruction with doubled semitendinosus and gracilis tendons and lateral extra-articular reconstruction.
      • Fox JA
      • Pierce M
      • Bojchuk J
      • et al.
      Revision anterior cruciate ligament reconstruction with nonirradiated fresh-frozen patellar tendon allograft.
      • Garofalo R
      • Djahangiri A
      • Siegrist O
      Revision anterior cruciate ligament reconstruction with quadriceps tendon-patellar bone autograft.
      • Grossman MG
      • ElAttrache NS
      • Shields CL
      • et al.
      Revision anterior cruciate ligament reconstruction: three- to nine-year follow-up.
      • Mayr HO
      • Willkomm D
      • Stoehr A
      • et al.
      Revision of anterior cruciate ligament reconstruction with patellar tendon allograft and autograft: 2- and 5-year results.
      ,
      • Salmon LJ
      • Pinczewski LA
      • Russell VJ
      • et al.
      Revision anterior cruciate ligament reconstruction with hamstring tendon autograft: 5- to 9-year follow-up.
      ,
      • Thomas NP
      • Kankate R
      • Wandless F
      • et al.
      Revision anterior cruciate ligament reconstruction using a 2-stage technique with bone grafting of the tibial tunnel.
      reported the postoperative radiographic outcomes of medial, lateral and patellofemoral compartments separately.

      Patients' characteristics

      The 19 papers included in the systematic review reported the results of 776 patients (61% males, 39% females) at a mean pooled follow-up of 5.8 years. Age at surgery was 30.2 years, while age at final follow-up was 36.0 years (table 1). A wide spectrum of both allografts and autografts was used for revision surgery; 81% of the patients were treated with an autograft and 19% with an allograft (figure 2). Reconstruction was performed mostly with a single-bundle one-stage technique; however, some studies reported a few cases of a two-stage procedure; only Thomas et al
      • Thomas NP
      • Kankate R
      • Wandless F
      • et al.
      Revision anterior cruciate ligament reconstruction using a 2-stage technique with bone grafting of the tibial tunnel.
      and Franceschi et al
      • Franceschi F
      • Papalia R
      • Del Buono A
      • et al.
      Two stage procedure in anterior cruciate ligament revision surgery: a five-year follow-up prospective study.
      reported the results of a two-stage procedure in all patients.
      Table 1Demographic details and information regarding graft choice and osteoarthritis grading scale of each study included in the systematic review
      AuthorsYearTreatedAt FUFollow-upAge at revisionGraft for revisionRadiographic score
      Wirth and Kohn199687878.0 years26 years57 (66%) ipsilateral BPTB autograftFairbanks
      (2.0–18.0)(14–41)30 (34%) ipsilateral quadriceps tendon autograft
      Fules et al200329294.2 years38±7 years26 (90%) ipsilateral hamstring autograftFairbanks
      (1.0–8.0)(24–53)2 (7%) ipsilateral BPTB autograft
      1 (3%) ipsilateral quadriceps tendon autograft
      Fox et al200438324.8 years28±8 years32 (100%) BPTB allograftIKDC
      (2.1–12–1)(16–57)
      Grossman et al200535295.6 years30.2 years22 (76%) BPTB allograftIKDC
      6 (21%) contralateral BPTB autograft
      1 (4%) Achilles tendon allograft
      Thomas et al200549496.2 years35.4 years38 (78%) ipsilateral hamstrings autograftJoint space narrowing
      (3–11)11 (22%) ipsilateral BPTB autograft
      Ferretti et al200630285.0 years34 years30 (100%) ipsilateral hamstrings autograftFairbanks
      (2–8)(21–39)
      Garofalo et al200631284.2 years27 years28 (100%) ipsilateral quadriceps tendon autograftIKDC
      (3.3–5.6)(18–41)
      Salmon et al200657507.4 years27 years26 (46%) contralateral hamstrings autograftIKDC
      (5.0–9.1)(15–39)30 (53%) ipsilateral hamstrings autograft
      1 (1%) hamstrings allograft
      Battaglia et al200795636.1 years31 years19 (30%) ipsilateral BPTB autograftFairbanks
      (3.0–13.2)(18–60)10 (16%) ipsilateral hamstrings autograft
      1 (2%) reharvested BPTB autograft
      1 (2%) contralateral BPTB autograft
      20 (32%) allograft
      Ahn et al200855554.0 years31.6 years21 (37%) ipsilateral hamstrings autograftFairbanks
      (1.8–10)(21–55)20 (36%) BPTB allograft
      15 (27%) Achilles allograft
      Diamantopoulos et al20081481076.0 years39±9 years45 (42%) ipsilateral hamstrings autograftJaeger and Wirth
      41 (38%) ipsilateral BPTB autograft
      21 (20%) ipsilateral quadriceps tendon autograft
      Ahn et al201150416.3 years32 years18 (44%) BPTB allograftFairbanks
      (3.5–9.8)(21–55)14 (34%) ipsilateral hamstrings autograft
      9 (22%) Achilles allograft
      Mayr et al201217155.7 years30±9 years17 (100%) BPTB allograftKellgren and Lawrence
      17145.8 years35±13 years17 (100%) ipsilateral BPTB autograftKellgren and Lawrence
      (2–6)5 (36%) ipsilateral BPTB autograft
      Franceschi et al201330306.8 years29.1 years30 (100%) ipsilateral hamstrings autograftFairbanks
      (5–9)(19–42)
      Gifstad et al201369567.5 years26.5 years22 (39%) contralateral BPTB autograftKellgren and Lawrence
      (2.8–13.2)25 (45%) ipsilateral hamstrings autograft
      9 (16%) ipsilateral BPTB autograft
      Griffith et al201318155.0 years27 years12 (80%) BPTB allograftKellgren and Lawrence
      (2.0–10.0)(18–57)1 (7%) ipsilateral hamstring autograft
      1 (7%) contralateral hamstring autograft
      1 (7%) ipsilateral BPTB autograft
      Kievit et al201330255.3 yearsNA12 (48%) posterior tibialis allograftIKDC
      (2.3–12.2)11(44%) Achilles allograft
      2 (8%) BPTB allograft
      Ventura et al201414144.2 yearsNA9 (64%) ipsilateral hamstrings autograftAhlbäck
      Dejour et al20151094.0 years30.3±4.4 years8 (89%) ipsilateral quadriceps tendon autograftAhlbäck
      (2.0–7.6)(26–37)1 (11%) ipsilateral hamstrings autograft
      BPTB, bone-patellar tendon-bone; FU, follow-up; NA, not applicable.
      Figure thumbnail gr2
      Figure 2Graft choice for revision anterior cruciate ligament (ACL) reconstruction of the studies included in the meta-analysis. Total number of included patients: 776.
      When the meniscal status was reported, a meniscal defect that resulted from previous meniscectomies was present in 25% for medial meniscus, 5% for lateral meniscus, 2% for both lateral and medial meniscus and 8% for not specified meniscus. A meniscal lesion was encountered at the time of revision ACL reconstruction in 40% of the patients for the medial meniscus, 30% for the lateral meniscus, 4% for the lateral and medial meniscus, and 5% for the not specified meniscus.
      Cartilage status was adequately reported in five studies. Normal cartilage status at the time of revision was reported in 50% of the cases. Concomitant procedures were reported in 7%, mostly represented by posterolateral corner (PLC) reconstruction (5%), medial collateral ligament (MCL) reconstruction (1%) and microfractures (1%).
      Radiographic outcomes were reported using different scales. Fairbanks classification was used in seven studies, showing 35% of the knees rated as normal, 43% as grade I, 14% as grade II, 8% as grade III and 0% as grade IV. IKDC 2000 score was used in six studies, reporting medial compartment graded as A in 55%, B in 33%, C in 9% and D in 3%, lateral compartment graded as A in 79%, B in 16%, C in 3% and D in 2%, patellofemoral compartment graded as A in 81%, B in 14%, C in 4% and D in 1%. Kellgren and Lawrence, Jaeger and Wirth and Ahlbäck grading were used in two studies each, while joint-space-narrowing grading was used in only one study.

      Postoperative OA prevalence

      Fifteen studies reported the radiographic outcomes of 627 patients at a mean 6.1 years after revision procedure using Fairbanks grading (7), Kellgren and Lawrence scale (3), IKDC score (2), Ahlbäck grading (2) or Jaeger and Wirth grading (1). The age at revision of the evaluated population was 31.4 years, while the age at final follow-up was 37.6 years.
      The incidence from the pooled data in the meta-analysis of the presence of OA after revision ACL reconstruction as determined by cut-off values was 61% (CI 50% to 70%; I2=79%, p<0.0001) (figure 3), ranging from 26%
      • Gifstad T
      • Drogset JO
      • Viset A
      • et al.
      Inferior results after revision ACL reconstructions: a comparison with primary ACL reconstructions.
      reported at 7.5 years follow-up using the Kellgren and Lawrence score, to 100%
      • Ventura A
      • Legnani C
      • Terzaghi C
      • et al.
      Revision surgery after failed ACL reconstruction with artificial ligaments: clinical, histologic and radiographic evaluation.
      reported after 4.2 years using Ahlbäck grading.
      Figure thumbnail gr3
      Figure 3Forest plot graphic of postoperative knee osteoarthritis (OA). In the horizontal axis, incidence in a 0–1 scale is reported. Larger squares represent a larger sample size.

      OA progression

      Nine studies reported preoperative and postoperative radiographic outcomes of 402 patients at a mean 6.3 years after the revision procedure, using Fairbanks grading (4), Ahlbäck grading (2), Kellgren and Lawrence scale (1), IKDC score (1) or Jaeger and Wirth grading (1). The age at revision of the evaluated population was 31.6 years, while the age at final follow-up was 38.0 years.
      The incidence from the pooled data in the meta-analysis of the preoperative presence of OA as determined by cut-off values was 41% (CI 26% to 57%; I2=82%, p<0.0001), while the postoperative incidence was 64% (CI 52% to 75%; I2=75%, p=0.0005). Therefore, progression from normal knee to OA was registered in 20% (CI 10% to 32%; I2=88%, p<0.0001) of the patients (figure 4). The OA progression ranged from 3% reported after 4 years with Fairbanks grading,
      • Ahn JH
      • Lee YS
      • Ha HC
      Comparison of revision surgery with primary anterior cruciate ligament reconstruction and outcome of revision surgery between different graft materials.
      to 41% reported after 6 years using Jaeger and Wirth grading.
      • Diamantopoulos AP
      • Lorbach O
      • Paessler HH
      Anterior cruciate ligament revision reconstruction: results in 107 patients.
      Ventura et al
      • Ventura A
      • Legnani C
      • Terzaghi C
      • et al.
      Revision surgery after failed ACL reconstruction with artificial ligaments: clinical, histologic and radiographic evaluation.
      reported 0% progression because all patients presented OA signs yet at preoperative status.
      Figure thumbnail gr4
      Figure 4Forest plot graphic of knee osteoarthritis (OA) at preoperative status (A), final follow-up (B) and progression between the two status (C). In the horizontal axis, incidence on a 0–1 scale is reported. Larger squares represent a larger sample size.

      Medial, lateral and patellofemoral OA

      Nine studies reported the postoperative OA prevalence for the medial, lateral and patellofemoral compartments of 297 patients at a mean follow-up of 5.8 years, using the IKDC score (5), Kellgren and Lawrence grading (2), Fairbanks grading (1) or joint space narrowing (1). The age at revision of the evaluated population was 30.7 years, while the age at final follow-up was 36.5 years.
      The pooled incidence from the meta-analysis of postoperative OA according to the cut-off values was 46% (CI 41% to 52%; I2=0%, p<0.7125) in the medial compartment, 23% (CI 13% to 36%; I2=81%, p<0.0001) in the lateral compartment and 27% (CI 18% to 38%; I2=74%, p=0.0002) in the patellofemoral compartment (figure 5).
      Figure thumbnail gr5
      Figure 5Forest plot graphic of postoperative knee osteoarthritis (OA) of the medial compartment (A), lateral compartment (B) and patellofemoral compartment (C). In the horizontal axis, incidence on a 0–1 scale is reported. Larger squares represent a larger sample size.

      Discussion

      The most important finding of this meta-analysis was that, despite revision ACL reconstruction after failed primary ACL, about 60% of patients presented OA signs at a mean follow-up of 6.2 years, while about 20% presented progression from normal knee to OA signs. These findings do not appear to portray an unexpected scenario, especially if preoperative knee status is considered. A recent meta-analysis of the clinical results of ACL revision reported 53% previous meniscectomies and 55% meniscectomies performed at the time of revision,
      • Wright RW
      • Gill CS
      • Chen L
      • et al.
      Outcome of revision anterior cruciate ligament reconstruction: a systematic review.
      which not substantially differs from the high percentages of meniscal defect or lesions reported in the present study. Also, cartilage status showed a similar trend, as nearly 50% of the knees presented abnormal cartilage at the time of revision in the same meta-analysis,
      • Wright RW
      • Gill CS
      • Chen L
      • et al.
      Outcome of revision anterior cruciate ligament reconstruction: a systematic review.
      in the multicentric Multicenter ACL Revision Study (MARS) cohort
      • Wright RW
      • Huston LJ
      • Spindler KP
      • MARS Group
      • et al.
      Descriptive epidemiology of the Multicenter ACL Revision Study (MARS) cohort.
      and in this study as well. Meniscal and cartilage lesions have been in fact correlated to a higher risk of developing knee OA,
      • Øiestad BE
      • Engebretsen L
      • Storheim K
      • et al.
      Knee osteoarthritis after anterior cruciate ligament injury: a systematic review.
      and therefore the high incidence of meniscal and cartilage lesions could motivate the OA incidence after revision ACL reconstruction reported in this study. Moreover several studies that evaluated the outcomes of ACL revision
      • Franceschi F
      • Papalia R
      • Del Buono A
      • et al.
      Two stage procedure in anterior cruciate ligament revision surgery: a five-year follow-up prospective study.
      • Griffith TB
      • Allen BJ
      • Levy BA
      • et al.
      Outcomes of repeat revision anterior cruciate ligament reconstruction.
      reported the worst radiographic results in patients with meniscal pathology and even in patients with late revision
      • Diamantopoulos AP
      • Lorbach O
      • Paessler HH
      Anterior cruciate ligament revision reconstruction: results in 107 patients.
      or with a longer period of knee instability,
      • Battaglia II, MJ
      • Cordasco FA
      • Hannafin JA
      • et al.
      Results of revision anterior cruciate ligament surgery.
      probably because these patients are more inclined to sustain further meniscal and cartilage injuries while with an unstable knee.
      The overall incidence of knee OA of 61% reported in this study appears to be higher than the 21–48% reported more than 10 years after ACL reconstruction.
      • Øiestad BE
      • Engebretsen L
      • Storheim K
      • et al.
      Knee osteoarthritis after anterior cruciate ligament injury: a systematic review.
      This higher incidence is also confirmed by several studies that compared the radiographic preoperative and postoperative results of revision ACL with a matched group of primary reconstruction,
      • Gifstad T
      • Drogset JO
      • Viset A
      • et al.
      Inferior results after revision ACL reconstructions: a comparison with primary ACL reconstructions.
      • Kievit AJ
      • Jonkers FJ
      • Barentsz JH
      • et al.
      A cross-sectional study comparing the rates of osteoarthritis, laxity, and quality of life in primary and revision anterior cruciate ligament reconstructions.
      • Won HH
      • Chang CB
      • Je MS
      • et al.
      Coronal limb alignment and indications for high tibial osteotomy in patients undergoing revision ACL reconstruction.
      which showed an increased percentage of patients with varus malalignment and medial OA already at preoperative status, and an almost double-fold incidence of OA changes after 2.8–8 years. Moreover, the same studies reported a higher incidence of meniscal and chondral lesions in patients who sustain revision ACL reconstruction. A similar finding was reported in the comparison between the MARS and Multicenter Orthopaedic Outcomes Network (MOON) cohorts of revision and primary reconstruction, respectively.
      • Borchers JR
      • Kaeding CC
      • Pedroza AD
      • et al.
      MOON Consortium and the MARS Group
      Intra-articular findings in primary and revision anterior cruciate ligament reconstruction surgery: a comparison of the MOON and MARS study groups.
      If compared with the normal population, in which the OA prevalence has been estimated at 19% after 45 years of age,
      • Suri P
      • Morgenroth DC
      • Hunter DJ
      Epidemiology of osteoarthritis and associated comorbidities.
      the scenario of the mid-30s patients who sustained revision surgery appears dramatic. Apart from meniscal and cartilage injuries, this could be also due to the high-level and prolonged sports activity of this subpopulation of patients. Also, the invasiveness of the revision ACL reconstruction compared with the primary ACL reconstruction (due to hardware removal and extensive debridement of the previous graft, especially when synthetic) could hypothetically be responsible for a certain level of OA changes. In fact, if single compartment OA is considered, it is noteworthy how the patellofemoral joint presented a higher or at least similar incidence of degeneration compared with the lateral tibiofemoral compartment. This represents a debated issue also after primary ACL reconstruction, where patellofemoral OA has been described as an under-recognised event.
      • Culvenor AG
      • Cook JL
      • Collins NJ
      • et al.
      Is patellofemoral joint osteoarthritis an under-recognised outcome of anterior cruciate ligament reconstruction? A narrative literature review.
      The cause of such a finding could be related to the frequent harvesting of bone-patellar tendon-bone (BPTB) or the quadriceps tendon during primary or revision reconstruction, as confirmed by the 43% and 39% of patellofemoral OA reported respectively by Mayr et al
      • Mayr HO
      • Willkomm D
      • Stoehr A
      • et al.
      Revision of anterior cruciate ligament reconstruction with patellar tendon allograft and autograft: 2- and 5-year results.
      and Thomas et al,
      • Thomas NP
      • Kankate R
      • Wandless F
      • et al.
      Revision anterior cruciate ligament reconstruction using a 2-stage technique with bone grafting of the tibial tunnel.
      that evaluated the mid-term results of patients with BPTB harvested during primary or revision surgery. Regarding the higher incidence of medial OA, it probably reflects the major involvement of the medial meniscus in traumatic or degenerative tears.
      However, since no controlled clinical trials are available with regard to OA progression after ACL revision, it was impossible to estimate the relative risk between the operated knee and the healthy or non-operatively treated knee. Therefore, the incidence of progression from the normal status to arthritic status of 20% reported in this study could both represent the natural history of such a multiple-injured knee, or even the inevitable natural consequence of the revision procedure. Conversely, it is impossible to figure out if the proportion of arthritic knees at final follow-up would have been higher without the correction of the instability. From the data available in this meta-analysis, it could be exclusively stated that, despite surgical correction, knee OA developed in one of the five patients. Therefore, the question “what would be the occurrence of OA be without intervention?” remains still open.
      This meta-analysis has several major limitations, which should caution the reader when interpreting the results. Above all, owing to the scarcity of clinical studies reporting radiographic outcomes of revision ACL, no stringent inclusion criteria were applied. Therefore, several different grading systems were used to evaluate knee OA based on cut-off values, thus dichotomising the knee status in normal and OA signs. This kind of evaluation, although already used by Øiestad et al,
      • Øiestad BE
      • Engebretsen L
      • Storheim K
      • et al.
      Knee osteoarthritis after anterior cruciate ligament injury: a systematic review.
      could have produced an inclusion bias, because the same knee could have been classified as normal or osteoarthritic depending on the grading scale utilised. Moreover, the cut-off system does not allow one to distinguish between mild, moderate or severe OA, even if in this study the evaluation of each single grading scale showed only small percentages of advanced OA. Also, the progression from mild-to-moderate or severe OA was not evaluated with such a system. On the other hand, the present method allowed the inclusion of more than 700 patients, strengthening the reported results. The heterogeneity of patient's age, follow-up, surgical techniques, graft choice and cartilage or meniscus status, even within each single study, did not allow one to stratify the outcomes of specific homogeneous subgroups, especially regarding age, graft choice and pre-existing cartilage damage or meniscal defects. However, it could be assumed that the reported results could be non-specifically generalised for any kind of revision reconstruction, because no exclusion criteria based on surgical technique or patient's characteristics were applied. Also, the lack of OA evaluation with long leg radiographs could represent a bias for this study, as the importance of lower limb alignment in the primary and also revision ACL reconstruction has been extensively underlined.
      • Dejour D
      • Saffarini M
      • Demey G
      • et al.
      Tibial slope correction combined with second revision ACL produces good knee stability and prevents graft rupture.
      • Won HH
      • Chang CB
      • Je MS
      • et al.
      Coronal limb alignment and indications for high tibial osteotomy in patients undergoing revision ACL reconstruction.
      • Bonin N
      • Ait Si Selmi T
      • Donell ST
      • et al.
      Anterior cruciate reconstruction combined with valgus upper tibial osteotomy: 12 years follow-up.
      ,
      • Sonnery-Cottet B
      • Mogos S
      • Thaunat M
      • et al.
      Proximal tibial anterior closing wedge osteotomy in repeat revision of anterior cruciate ligament reconstruction.
      ,
      • Zaffagnini S
      • Bonanzinga T
      • Grassi A
      • et al.
      Combined ACL reconstruction and closing-wedge HTO for varus angulated ACL-deficient knees.
      Lastly, the choice to consider 4 years as the minimum follow-up required for study inclusion reflects the paucity of the available literature, and could be considered a compromise between the necessity to obtain a sufficient number of studies and the minimum period to detect noticeable OA signs or OA progression.

      Conclusions

      Knee OA was present in almost 60% of the patients at a mean 6.2 years follow-up after ACL revision, with an almost double-fold incidence in the medial compartment OA relative to the lateral and the patellofemoral compartments. Moreover, progression from normal status to OA signs was detected in about 20% of the cases. These findings highlight OA as a concrete problem after revision ACL surgery, which could affect clinical results, sport participation and daily life activities. Therefore, early reconstruction, minimally invasive techniques, sparing of bone stock and the use of allografts or contralateral grafts could represent possible solutions to reduce joint damage and hypothetically oppose the onset or progression of knee OA. Moreover, the treatment of cartilage pathology and the preservation or substitution of damaged menisci should also be encouraged.

      References

        • Ajuied A
        • Wong F
        • Smith C
        • et al.
        Anterior cruciate ligament injury and radiologic progression of knee osteoarthritis: a systematic review and meta-analysis.
        Am J Sports Med. 2014; 42: 2242-2252
        • Fules PJ
        • Madhav RT
        • Goddard RK
        • et al.
        Revision anterior cruciate ligament reconstruction using autografts with a polyester fixation device.
        Knee. 2003; 10: 335-340
        • Gifstad T
        • Drogset JO
        • Viset A
        • et al.
        Inferior results after revision ACL reconstructions: a comparison with primary ACL reconstructions.
        Knee Surg Sports Traumatol Arthrosc. 2013; 21: 2011-2018
        • Dejour D
        • Saffarini M
        • Demey G
        • et al.
        Tibial slope correction combined with second revision ACL produces good knee stability and prevents graft rupture.
        Knee Surg Sports Traumatol Arthrosc. 2015; 23: 2846-2852
        • Ventura A
        • Legnani C
        • Terzaghi C
        • et al.
        Revision surgery after failed ACL reconstruction with artificial ligaments: clinical, histologic and radiographic evaluation.
        Eur J Orthop Surg Traumatol. 2014; 24: 93-98
        • Øiestad BE
        • Engebretsen L
        • Storheim K
        • et al.
        Knee osteoarthritis after anterior cruciate ligament injury: a systematic review.
        Am J Sports Med. 2009; 37: 1434-1443
        • Hedges LV
        • Vevea JL
        Fixed-and random-effects models in meta-analysis.
        Psychol Methods. 1998; 3: 486-504
        • Higgins JP
        • Thompson SG
        • Deeks JJ
        • et al.
        Measuring inconsistency in metaanalyses.
        BMJ. 2003; 327: 557-560
        • Ahn JH
        • Lee YS
        • Chang MJ
        • et al.
        Analysis of revision anterior cruciate ligament reconstruction according to the combined injury, degenerative change, and MRI findings.
        Knee. 2011; 8: 382-386
        • Ahn JH
        • Lee YS
        • Ha HC
        Comparison of revision surgery with primary anterior cruciate ligament reconstruction and outcome of revision surgery between different graft materials.
        Am J Sports Med. 2008; 36: 1889-1895
        • Battaglia II, MJ
        • Cordasco FA
        • Hannafin JA
        • et al.
        Results of revision anterior cruciate ligament surgery.
        Am J Sports Med. 2007; 35: 2057-2066
        • Diamantopoulos AP
        • Lorbach O
        • Paessler HH
        Anterior cruciate ligament revision reconstruction: results in 107 patients.
        Am J Sports Med. 2008; 36: 851-860
        • Ferretti A
        • Conteduca F
        • Monaco E
        • et al.
        Revision anterior cruciate ligament reconstruction with doubled semitendinosus and gracilis tendons and lateral extra-articular reconstruction.
        J Bone Joint Surg Am. 2006; 88: 2373-2379
        • Fox JA
        • Pierce M
        • Bojchuk J
        • et al.
        Revision anterior cruciate ligament reconstruction with nonirradiated fresh-frozen patellar tendon allograft.
        Arthroscopy. 2004; 20: 787-794
        • Franceschi F
        • Papalia R
        • Del Buono A
        • et al.
        Two stage procedure in anterior cruciate ligament revision surgery: a five-year follow-up prospective study.
        Int Orthop. 2013; 37: 1369-1374
        • Garofalo R
        • Djahangiri A
        • Siegrist O
        Revision anterior cruciate ligament reconstruction with quadriceps tendon-patellar bone autograft.
        Arthroscopy. 2006; 22: 205-214
        • Griffith TB
        • Allen BJ
        • Levy BA
        • et al.
        Outcomes of repeat revision anterior cruciate ligament reconstruction.
        Am J Sports Med. 2013; 41: 1296-1301
        • Grossman MG
        • ElAttrache NS
        • Shields CL
        • et al.
        Revision anterior cruciate ligament reconstruction: three- to nine-year follow-up.
        Arthroscopy. 2005; 21: 418-423
        • Kievit AJ
        • Jonkers FJ
        • Barentsz JH
        • et al.
        A cross-sectional study comparing the rates of osteoarthritis, laxity, and quality of life in primary and revision anterior cruciate ligament reconstructions.
        Arthroscopy. 2013; 29: 898-905
        • Mayr HO
        • Willkomm D
        • Stoehr A
        • et al.
        Revision of anterior cruciate ligament reconstruction with patellar tendon allograft and autograft: 2- and 5-year results.
        Arch Orthop Trauma Surg. 2012; 132: 867-874
        • Salmon LJ
        • Pinczewski LA
        • Russell VJ
        • et al.
        Revision anterior cruciate ligament reconstruction with hamstring tendon autograft: 5- to 9-year follow-up.
        Am J Sports Med. 2006; 34: 1604-1614
        • Thomas NP
        • Kankate R
        • Wandless F
        • et al.
        Revision anterior cruciate ligament reconstruction using a 2-stage technique with bone grafting of the tibial tunnel.
        Am J Sports Med. 2005; 33: 1701-1709
        • Wirth CJ
        • Kohn D
        Revision anterior cruciate ligament surgery: experience from Germany.
        Clin Orthop Relat Res. 1996; 325: 110-115
        • Wright RW
        • Gill CS
        • Chen L
        • et al.
        Outcome of revision anterior cruciate ligament reconstruction: a systematic review.
        J Bone Joint Surg Am. 2012; 94: 531-536
        • Wright RW
        • Huston LJ
        • Spindler KP
        • MARS Group
        • et al.
        Descriptive epidemiology of the Multicenter ACL Revision Study (MARS) cohort.
        Am J Sports Med. 2010; 38: 1979-1986
        • Won HH
        • Chang CB
        • Je MS
        • et al.
        Coronal limb alignment and indications for high tibial osteotomy in patients undergoing revision ACL reconstruction.
        Clin Orthop Relat Res. 2013; 471: 3504-3511
        • Borchers JR
        • Kaeding CC
        • Pedroza AD
        • et al.
        • MOON Consortium and the MARS Group
        Intra-articular findings in primary and revision anterior cruciate ligament reconstruction surgery: a comparison of the MOON and MARS study groups.
        Am J Sports Med. 2011; 39: 1889-1893
        • Suri P
        • Morgenroth DC
        • Hunter DJ
        Epidemiology of osteoarthritis and associated comorbidities.
        PM R. 2012; 4: S10-S19
        • Culvenor AG
        • Cook JL
        • Collins NJ
        • et al.
        Is patellofemoral joint osteoarthritis an under-recognised outcome of anterior cruciate ligament reconstruction? A narrative literature review.
        Br J Sports Med. 2013; 47: 66-70
        • Bonin N
        • Ait Si Selmi T
        • Donell ST
        • et al.
        Anterior cruciate reconstruction combined with valgus upper tibial osteotomy: 12 years follow-up.
        Knee. 2004; 11: 431-437
        • Sonnery-Cottet B
        • Mogos S
        • Thaunat M
        • et al.
        Proximal tibial anterior closing wedge osteotomy in repeat revision of anterior cruciate ligament reconstruction.
        Am J Sports Med. 2014; 42: 1873-1880
        • Zaffagnini S
        • Bonanzinga T
        • Grassi A
        • et al.
        Combined ACL reconstruction and closing-wedge HTO for varus angulated ACL-deficient knees.
        Knee Surg Sports Traumatol Arthrosc. 2013; 21: 934-941