Outcomes of cartilage repair techniques for chondral injury in the hip—a systematic review

The aim of the study was to assess the options of treatment and their related outcomes for chondral injuries in the hip based on the available evidence whilst highlighting new and innovative techniques. A systematic review of the literature from PubMed (Medline), EMBASE, Google Scholar, British Nursing Index (BNI), Cumulative Index to Nursing and Allied Health Literature (CINAHL) and Allied and Complementary Medicine Database (AMED) was undertaken from their inception to March 2017 using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Clinical outcome studies, prospective/retrospective case series and case reports that described the outcome of cartilage repair technique for the chondral injury in the hip were included. Studies on total hip replacement, animal studies, basic studies, trial protocols and review articles were excluded. The systematic review found 21 relevant papers with 596 hips. Over 80% of the included studies were published in or after 2010. Most studies were case series or case reports (18 studies, 85.7%). Arthroscopy was used in 11 studies (52.4%). The minimum follow-up period was six months. Mean age of the participants was 37.2 years; 93.5% of patients had cartilage injuries of the acetabulum and 6.5% of them had injuries of the femoral head. Amongst the 11 techniques described in the systematic review, autologous matrix-induced chondrogenesis, osteochondral autograft transplantation and microfracture were the three frequently reported techniques. Over ten different techniques are available for cartilage repair in the hip, and most of them have good short- to medium-term outcomes. However, there are no robust comparative studies to assess superiority of one technique over another, and further research is required in this arena.


Introduction
Isolated chondral and osteochondral defects within the hip joint often present a technical challenge for the hip surgeon. Common causes of cartilage damage in the hip include femoroacetabular impingement (FAI), developmental dysplasia, osteonecrosis, osteochondritis dissecans, loose bodies, slipped capital femoral epiphysis, and trauma [1][2][3][4][5]. Amongst them, FAI has increasingly gained recognition as a major cause of chondral injury and subsequent development of arthritis in the hip joint [6][7][8][9][10]. In CAM FAI, the abnormal contact between the aspherical femoral head-neck junction and the acetabular rim results in a large amount of shear stress being transmitted to the labro-chondral junction. Over a period of time, labral detachment and acetabular chondral damage ensues [2,11,12]. On the other hand, the pincer FAI, in which a deep or retroverted acetabulum makes contact with a normal-shaped femoral neck, has a recognised pattern of damage to the labrum, femoral head cartilage and a postero-medial acetabular countercoup lesion [13]. Furthermore, in imaging and surgical techniques like hip arthroscopy have led to increased recognition of chondral lesions. The incidence of chondral lesions at hip arthroscopy for FAI has been reported to be up to 67.3% of the patients in one series [14].
There is relatively little information about articular cartilage restoration in the hip when compared with what is known about cartilage restoration in the knee. Currently, most cartilage repair methods for the hip are based on basic science and strategies that were developed for the knee. Awareness of young adult hip disease has been increasing in recent years, and thus, the field of hip preservation continues to develop; several new innovative techniques have been performed and described in the literature. They include microfracture, autologous chondrocyte implantation (ACI), matrix-associated chondrocyte implantation (MACI), autologous matrixinduced chondrogenesis (AMIC), osteochondral autograft/ allograft transplantation, implantation of artificial plug, sticking down of chondral flaps with fibrin adhesive and an intraarticular injection of bone marrow mesenchymal stem cells (BM-MSCs).
Currently, there is a gap in information particularly regarding systematic reviews in the literature that provide hip surgeons with evidence-based recommendations, therefore, on treating cartilage injuries in the hip. The aim of this study was to provide the reader with options of treatment and their related outcomes for chondral injuries in the hip based on the available evidence whilst highlighting new and innovative techniques involved in chondral repair.

Search strategy
Two reviewers (NN and CG) searched the online databases (PubMed (Medline), EMBASE, Google Scholar, British Nursing Index (BNI), Cumulative Index to Nursing and Allied Health Literature (CINAHL) and Allied and Complementary Medicine Database (AMED) for literature describing the outcome of cartilage repair techniques for the chondral injury in the hip. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were used for designing this study. A detailed search strategy is described in the Appendix.

Study screening/data abstraction
The inclusion and exclusion criteria are shown in Table 1. Both the reviewers independently abstracted the relevant study data from the final pool of included articles and recorded this data on a spreadsheet designed a priori. Participant-specific demographics extracted from each study included the number of hips, gender distribution, mean age with range (years), length of follow-up, location of the cartilage injury (acetabulum or femoral head), surgical approach (open dislocation, arthroscopy or injection), cartilage restoration technique used in the study, pre-operative condition of the damaged cartilage, final outcome and specific comments (if any).

Statistics
The abstracted evidence was collected and analysed using Microsoft Excel 2013 spread sheet. Statistical analysis in this study focused on descriptive statistics.
Details of the 21 studies included are shown in Table 3.

Discussion
Our objective was to discuss the outcomes of the current strategies for restoration of focal chondral injuries in the hip. This study reviews all the cases of cartilage repair for the chondral injuries in the hip (596 cases) reported in the English literature and describes the outcomes of 11 techniques (including debridement). Cartilage injuries in the hip have been previously shown to result in poor long-term outcomes, including pain and early secondary degenerative change followed by the subsequent development of osteoarthritis [15,16]. The current trend is to focused on treating isolated cartilage damage and underlying morphological pathology in younger patients in order to prevent progression to end-stage degeneration. Although a number of procedures for the management of chondral lesions in other large joints (e.g. knee) have been reported, there currently remains little information available for appropriate management of these lesions in the hip [17]. All the techniques found in the systematic review are described and discussed below.

Debridement
Debridement of a cartilage flap from a chondral injury may allow symptoms to resolve and permit a return to activity or sports [6,18]. Arthroscopy is essential for the diagnosis of an unstable flap if pre-operative imaging is unclear, and arthroscopic debridement is often the definitive therapy. Fontana et al.. [19] carried out a controlled retrospective study of 30 patients (15 ACI, 15 arthroscopic debridements) affected by a post-traumatic hip chondropathy (Outerbridge classification grades 3-4, measuring 2 cm 2 in area or more). The postoperative Harris Hip Scores (HHS) in the ACI group were significantly better than those in the debridement group.

Microfracture
Microfracture involves the use of an arthroscopic awl or drill to perforate exposed subchondral bone to create multiple holes and provide an entry portal for marrow-derived cells. The rationale of the technique is to recruit mesenchymal stem cells into the cartilage defect to create fibrocartilage. Following microfracture, a marrow clot forms and provides the ideal environment for mesenchymal stem cells to differentiate into stable repair tissue [20]. The advantages of this technique are that it is technically straightforward, can be performed arthroscopically, without donor site morbidity, and has a low cost. The disadvantage compared with other cartilage repair techniques is that it produces less type II cartilage and has different biomechanical properties than hyaline cartilage, which may raise questions of its resilience and longevity     Walking with 2 crutches and weight bearing as tolerated was allowed on the first post-operative day. The median pre-operative mHHS, WOMAC and VAIL scores were 64.3, 73 and 56.5, respectively, and they increased to 91, 97 and 83 at final follow-up (p < 0.05). The VAS score also improved from a median of 6 to 2. Four patients received a THR (13% of the hips) at the median of 9 months post-intervention (range 6-36 months); 80 mL of bone marrow was aspirated from the anterior iliac crest during hip arthroscopy. Each patient received The Tegner-Lysholm score at latest follow-up ranged from 5 to 9 (mean, 7.4). All patients . and radiographic confirmation of trochanteric union, progressive weight bearing was encouraged. were able to return to their pre-operative level of function with the exception of patient 6 whose contralateral hip precluded participation. There was no obvious asymmetric joint space narrowing visible on an AP pelvis view in any of the patients.

2012
Severe osteochondral lesion with a subchondral cyst on the femoral head seen on MRI.
Restricted weight bearing during 4 weeks (walking with crutches and plantar touch  ).
Patients were kept non-weight bearing for 6 weeks and then progressed to weight bearing as tolerated. HHS increased from 52.8    There was also a full-thickness cartilaginous injury at the apex of the fracture, in the anterior-superior weight-bearing zone of the femoral head. This zone of injury was approximately 10 mm in size.
Post-operatively, the patients were kept non-weight bearing for 6 weeks and then progressed to weight bearing as tolerated. This intact 1-cm delaminated articular cartilage flap (Outerbridge grade 0) was partially off the subchondral bone.
Patient was allowed 30% weight bearing with crutches for 6 weeks, gradually progressing to 100% over the following 2 weeks. The patient reported being pain-free 90% of the time with pain 2/10 at worst. He scored 96 on mHHS, 93 on HOS Activities of Daily Living subscale and 81 on HOS Sports subscale.

2008
The average acetabular chondral lesion size was 163 mm 2 . All lesions were located in the superior acetabular quadrant.
Weight bearing was restricted to toe-touch for 8 weeks.
The average percent fill of the acetabular chondral lesions at second look was 91% (range, 25 to 100%). Eight of the 9 patients had grade 1 or 2 repair product at second look (grade 1 was normal-appearing articular cartilage, difficult to discern borders of lesion and normal surrounding cartilage; grade 2 was mild fibrillation, discoloured, softer-than-normal cartilage; grade 3 was deep fissures or cobblestone surface, no exposed Other comments bone; and grade 4 was full-thickness cartilage loss with exposed subchondral bone). One patient who had diffuse osteoarthritis failed, with only 25% coverage with a grade 4 appearance of the repair product 10 months after index arthroscopy and required total hip arthroplasty 66 months after the index microfracture.
2008 MR arthrogram revealed full-thickness loss of the surrounding articular cartilage on the major weight-bearing portion of the femoral head. Intra-operative measurement of the chondral defect measured 4.0 cm by 2.5 cm around intact osteochondral plugs.
Progressive weight-bearing activity can begin as early as 6 weeks but is usually delayed until 8 to 12 weeks.
Two years later, the patient remains free of pain, her post-operative contrast-enhanced MRI demonstrates repair tissue fill and radiographs showed a normal joint space. The patient had progression of disease after prior autologous osteochondral mosaicplasty. Fixation of the membrane was performed with the use of 6.0 Vicryl suture. Fibrin glue was used to further seal the membrane.

2003
The diameter of the round defect was 14 mm, and its depth was 16 mm.
Partial weight bearing was permitted at 6 weeks and full weight bearing at 10 weeks after the surgery.
HHS improved from 69 to 100 points. At 6 months post-operatively, the patient showed the full range of painless motion with no further complaints of rest pain or pain related to activities.  [20]. Also, the concentration of mesenchymal cells in the bone marrow is relatively low and their chondrogenic potential declines with age [21]. Philippon et al. [22] reported the outcome of microfracture in nine patients with a full-thickness chondral defect of the acetabulum. The average percent fill of the acetabular chondral lesions at second-look arthroscopy was 91%, and eight of the nine patients had grade 1/2 repair product at second look. Karthikeyan et al. [23] described the outcome of 20 patients who underwent arthroscopic surgery for FAI with a localised full-thickness acetabular chondral defect which were treated by microfracture. At an average follow-up of 17 months, 19 of the 20 patients had a mean fill of 96% with macroscopically good quality repair tissue. Zaltz and Leunig [24] reviewed ten patients with symptomatic FAI diagnosed with parafoveal chondral defects confirmed at the time of surgical dislocation. Seven of the ten patients were treated by microfracture (other 3 were treated by AMIC), and all the patients were able to return to their pre-operative level of function with the exception of one patient who had a problem in the contralateral hip. At the last follow-up, there was no obvious asymmetric joint space narrowing visible on an AP pelvis view in any of the patients. Fontana et al. [25] compared the outcome of 77 patients who had a microfracture and 70 patient who had AMIC for cartilage injuries in the hip. Although the outcome in both groups significantly improved at six months and one year post-operatively, the outcome in the microfracture group slowly deteriorated four years after surgery, whilst that in the AMIC group remained stable.

Autologous chondrocyte implantation
ACI includes the harvest of chondrocytes with growth and expansion at an off-site facility, followed by reimplantation of the cells into the affected area. ACI is indicated for symptomatic, well-contained defects that are between 2 and 10 cm 2 and with less than 6-8 mm of bone loss [26]. Most surgeons who perform ACI regularly are now using a synthetic collagen membrane to cover the implanted chondrocytes [19,27]. Ellender and Minas [27] presented a clinical case and described ACI for a femoral head chondral defect of 10 cm 2 in a 19-year-old female college student who had progression of disease after prior mosaicplasty. Two years after ACI, she remained free of pain. Her post-operative contrast-enhanced MRI demonstrated repair tissue fill and radiographs showed a normal joint space without any sign of change.

Matrix-associated chondrocyte implantation
MACI is a second-generation ACI technique that utilises absorbable scaffolds to support the implanted chondrocytes during healing. Theoretically, it should restore hyaline cartilage at the defect. Unfortunately, same as ACI, it is a two-stage procedure where chondrocytes are harvested from the patient, cultured and then returned to the patient via open surgical dislocation of the hip which is a technically demanding surgical approach. Mancini and Fontana [28] assessed and compared the clinical outcomes of arthroscopic MACI and AMIC for the treatment of acetabular chondral defects between 2 and 4 cm 2 consequent to FAI. In both groups, significant improvement in modified HHS (mHHS) was measured over baseline levels at six months post-operation. It continued to improve up to three years post-operation and remained stable until five years follow-up. There was no statistically significant difference between the two groups.

Autologous matrix-induced chondrogenesis
AMIC is a novel single-step procedure in which the microfracture technique has been enhanced by the use of a collagen matrix. The Chondro-Gide matrix is placed in the defect and a porcine collagen I/III matrix is sewn over the lesion to stabilise the fragile blood clot that arises from the microfracture to provide a stable infrastructure for the formation of repair tissue [29]. No cells have to be harvested, cultured and re-implanted in AMIC. Therefore, there is no harvest site morbidity, and the operation can be performed as a single procedure. Moreover, AMIC does not require complex cell expansion techniques. Other than comparative studies with microfracture [24,25] or MACI [28] described above, Leunig et al. [30] reported six patients with AMIC using surgical dislocation of the hip. No complications occurred, and good post-operative outcome scores were reported. Fontana [31] treated 201 patients with AMIC arthroscopically for Outerbridge grade III/IV chondral lesions of the acetabulum. Modified HHS improved significantly at six months postoperatively in comparison with pre-operative levels, reaching the highest level of improvement at the three year follow-up.

Osteochondral autograft transplantation (mosaicplasty)
Mosaicplasty involves transplanting healthy, mature cartilage from a non-weight-bearing part of the hip or knee to an articular defect. The transplanted cartilage integrates with the adjacent host cartilage via fibrocartilage [32]. The inferior aspect of the femoral head, the femoral head-neck junction and the periphery of the femoral trochlea of the knee can be the potential donor sites. Mosaicplasty offers many potential advantages, including the ability to transfer new mature hyaline cartilage into the defect in a single-stage procedure and the absence of potential disease transmission, which can occur in allograft transplantation. On the contrary, owing to the autologous nature of this technique, it is limited by donor site morbidity, graft availability and the potential for dead space between the grafts [32]. Hart et al. [33] first reported the case of an osteochondral defect of the femoral head and subsequent treatment using mosaicplasty with open surgical dislocation of the hip. At six months following surgery, the patient had a full range of painless movement of the hip with no further complaints of pain related to activities. Emre et al. [34] also presented a case where the defect of the femoral head was treated with surgical dislocation of the hip and mosaicplasty. The patient was symptom-free with nearly complete incorporation of the graft radiologically at three years after the operation. Nam et al. [35] reported two cases of a chondral defect on the femoral head after a traumatic hip dislocation, treated with mosaicplasty from the ipsilateral knee, and the inferior femoral head, respectively. At 1 and five years of follow-up, MRI showed good autograft incorporation with the maintenance of articular surface conformity. Krych et al. [36] reported two cases of post-traumatic osteochondral defects of the femoral head. Both the patients were treated with mosaicplasty from the ipsilateral knee to the femoral head, with successful clinical and radiological results at a mean follow-up of 4.3 years. Girard et al. [37] treated 10 patients for femoral cartilage defects by mosaicplasty of the femoral head through a trochanteric flip osteotomy with surgical dislocation of the hip. At the mean follow-up of 29.2 months, clinical score and range of motion improved significantly. All radiological investigations at the latest follow-up showed that the grafts were wellincorporated at the site of mosaicplasty with intact cartilage over them and smooth interfaces between articulating bony surfaces.

Osteochondral allograft transplantation
Mosaicplasty has been shown to be a useful procedure, but there can be donor site morbidity and there is a limit to the size of the treatable defect. Allograft transplantation can also be a successful solution for the treatment of cartilage defects. It offers not only the potential advantages of transferring immediate functional hyaline cartilage but also the ability to resurface a large area without associated donor site morbidity. Potential allograft donor sources for defects within the acetabular side of the hip were a cadaveric acetabulum or medial tibial plateau. Cartilage is relatively immunoprivileged and avascular; thus, the host immune reaction is considered to be limited [38]. Allograft bone becomes necrotic and is reabsorbed via creeping substitution during the healing process. This provides a scaffold and supports the articular surface as part of gradual incorporation [39]. In the systematic review, Krych et al. [40] reported their experience in two patients who underwent osteochondral allograft transplantation for the acetabular cartilage defects. MRI at 18 months in both cases demonstrated incorporation of the graft into the host acetabulum. Hip Outcome Scores (HOS) were 100 points each in both patients two years post-operatively.

Direct cartilage suture repair
Delamination is a full-thickness cartilage separation from the underlying subchondral bone, which forms an unstable flap at risk for complete detachment [41]. Our review found a case report that presented direct cartilage repair as a possible technique to treat large delaminated full-thickness acetabular cartilage repairs. Sekiya et al. [42] described a case of a 17-year-old boy presented with bilateral hip pain because of bilateral CAMtype FAI and a 1-cm delaminated unstable cartilage flap in the anterior-superior acetabulum. Arthroscopic microfracture underneath the flap of anterior-superior acetabular cartilage and an absorbable monofilament suture repair of the cartilage was conducted. At two years post-operatively, the patient reported 95% of normal function for both hips. Overall, the patient was satisfied with the outcome including a score of 96 on the mHHS, 93 on the HOS Activities of Daily Living subscale and 81 on the HOS Sports subscale at the final follow-up.

Fibrin adhesive
The earliest stage in the formation of an articular cartilage flap is delamination of the overlying articular cartilage from the underlying subchondral bone [43]. Particularly, if the articular cartilage itself may contain a significant number of viable chondrocytes, debriding such an area of chondral instability seems an unnecessary surgical procedure. Fibrin adhesive is a biological substance, which has already been used in general surgery, ophthalmology, neurosurgery, otolaryngology and orthopaedics, thanks to its adhesive properties [44][45][46][47][48]. This procedure involved creating an incision at the periphery of the acetabular labrum and passing an awl underneath to create microfracture. Fibrin glue was inserted between subchondral bone and delaminated cartilage, and the cartilage was pressed down until the adhesive had set. Tzaveas and Villar [49] analysed the efficacy of using fibrin adhesive for arthroscopic repair of chondral delamination lesions with intact gross cartilage structure in 19 patients. Mean mHHS was improved significantly after surgery, and in all five patients who underwent revision arthroscopy at a later date, the chondral repair appeared intact. Stafford et al. [50] reported the results of 43 patients with FAI who have undergone fibrin adhesive technique for reattachment of delaminated chondral flaps. Both mHHS for pain and function improved significantly after the operation. In three patients who required further arthroscopic interventions for persistent symptoms created by iliopsoas irritation, the previously repaired articular cartilage was found in a good condition.

Intra-articular BM-MSC injection
Adult MSCs were originally believed to only differentiate into tissue-specific cells. However, these cells were recently proven to have the ability to differentiate into a different tissue in response to specific signals released by the site of injury, including cartilage injury [51,52]. Adding to animal studies, several authors reported on intra-articular injection of MSCs into the knee for the treatment of cartilage defects and showed good results with regard to pain and clinical outcomes [53][54][55][56]. Injected MSCs were incorporated into the articular cartilage of the injected joint. They integrate into the surface of the cartilage and also the interior of the cartilage [52]. Mardones et al. [57] first reported the outcome of intra-articular BM-MSC injection for the cartilage injury in the hip. Three intra-articular injections of 20 × 10 6 BM-MSCs were conducted from four to six weeks post-operatively in 29 hips that received hip arthroscopy for FAI and focal cartilage injuries. Clinical outcome scores and VAS improved significantly after surgery, and no major complications had been reported at the time of the last follow-up.

Artificial plug
The systematic review found two articles that used an artificial plug, and both of them utilise the TruFit cartilage/bone (CB) plug (Smith & Nephew). It is a resorbable polymer scaffold that can be inserted into osteochondral defects, which acts as a scaffold that provides structural support. Also, native marrow elements can migrate into the plug to promote bone in-growth as well as articular cartilage regeneration. Field et al. [58] described the use of TruFit for the treatment of acetabular cystic cartilage lesions in four patients. Patients underwent hip arthroscopy followed by the antegrade insertion of a plug through the ilium until the surface of the plug coincided with the articular surface. At ten months follow-up, patients reported increased function and improvement in Non-Arthritic Hip Score (NAHS). CT and MRI showed incorporation and continued healing of the plug six months post-operatively. Vundelinckx et al. [59] reported a case of a 34-year-old employee (gender was not described) who underwent TruFit for an osteochondral injury of the femoral head. MRI at six months showed the TruFit was placed in situ whilst there was an irregularity on the border of the articular cartilage surface. They mentioned it was very difficult to interpret early MRI images of ingrowth of TruFit plugs, as described by authors of past radiographic studies [60].
Of the 21 studies found in the systematic review, only 3 studies are level IIIb (retrospective comparative study) and the rest were level IV (case series/report). Two studies described superiority of one cartilage repair method over another [19,25], and one study showed there was no difference in clinical outcome between two methods [28]. Fontana's study [19] was limited by the reduced number of patients and the lack of an objective method for the evaluation of the results. Other limitations are the criteria for patient inclusion and selection bias in the randomisation process. Fontana's study [25] and Mancini's study [28] were also limited by the lack of randomisation, and clinical outcomes were only assessed using the mHHS.
The strengths of this systematic review include the pursuit of knowledge in an important novel area of investigation and a rigorous methodological approach. Regarding the methodological approach, a broad-based and comprehensive literature search of multiple databases with multiple reviewers allowed for a very inclusive approach to capture the vast majority of existing literature. Nonetheless, there are limitations which include the inclusion of English only studies and the overall low level of evidence available in the included studies on this topic (mostly level IIIb and IV studies). Retrospective designs are prone to data inaccuracy as well as missing information, which subject them to selection and detection bias. Without a doubt, this diminishes the accuracy of the data collected and, therefore, limits the quality of a systematic review, whilst this current level of evidence reflects the novel and emerging nature of cartilage repair strategies in the hip joint. Additionally, our results include a wide spectrum of pathologies and methods of treatment, which also made drawing conclusions and giving specific guidelines difficult. Furthermore, pre-operative condition and post-operative rehabilitation protocol were different in each study, which made comparison among studies difficult as well. Future studies should address comparative effectiveness of the various treatment options, and long-term registry-based studies that report patient reported outcomes and radiographic outcomes will help inform treatment decisions.

Conclusion
Although there are many different cartilage restoration techniques available, current best evidence does not support any one surgical technique as a superior method for treating cartilage injuries in the hip. Unfortunately there remains a paucity of randomised trials with long-term follow-up, which makes it difficult to perform a meaningful assessment of the outcome of each procedure. Of the 21 studies found in the systematic review, AMIC, mosaicplasty and microfracture were relatively well-reported, though they were only described in very limited case series. Also, only two studies described superiority of one cartilage repair method over another-one showed superiority of AMIC over microfracture [25] and another showed superiority of ACI over debridement [19], and one study showed that there was no statistically significant difference between MACI and AMIC in terms of post-operative mHHS [28]. To make any specific recommendations for orthopaedic surgeons with regards to treatment decisions, adequately powered long-term large-scale high-quality randomised-control trials focusing on two or three specific methods of treatment need to be conducted in the future.
Contribution of authors VK takes responsibility for the integrity of the work as a whole, from inception to the finished manuscript. NN, CG, AD, OA and VK were responsible for the conception and design; NN, CG and VK for the collection, assembly, analysis and interpretation of data; NN, OA and VK for drafting; and NN, CG, AD, OA and VK for the final approval of the manuscript and for the critical revision for important intellectual contents.

Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of interest.
Ethical approval This article does not contain any studies with human participants.