No evidence exists on outcomes of non-operative management in patients with femoroacetabular impingement and concomitant Tönnis Grade 2 or more hip osteoarthritis: a scoping review

Purpose The purpose of this scoping review was to assess the outcomes of all the non-operative modalities of management for femoroacetabular impingement (FAI) and concomitant osteoarthritis (OA) Tönnis Grade 2 or more. Methods A systematic search of PubMed was performed from inception to December 1st 2021 for literature on outcomes of non-operative management strategies for young adults with symptomatic FAI using the PRISMA Extension for Scoping Reviews guidelines. Cohorts investigating FAI and concomitant hip OA Tönnis Grade 2 or more were considered eligible. Studies not written in English or German, below level 4 evidence, and reviews were excluded. A secondary analysis for FAI without OA stratification was conducted after the initial screening to allow identification of available non-operative interventions. Results No study reported outcomes separately for non-operative management of FAI with Tönnis Grade 2 OA or more and as such, did not fulfil the inclusion criteria. A secondary analysis included 24 studies that reported on outcomes for non-operative interventions for FAI irrespective of the degree of degeneration. Three studies investigated the efficacy of hyaluronic acid injection, 5 reports investigated corticosteroid injections, 2 studies evaluated the outcomes of hip bracing and 16 studies included a physiotherapy programme. Associations between the aforementioned interventions were analysed. There is level I evidence supporting the efficacy of activity modification and hip-specific physiotherapy for FAI and mild OA. Core-strengthening exercises are prevalent amongst successful regimens in the literature. Contradictory evidence questions the efficacy of hip bracing even for short-term outcomes. Corticosteroid injections have mostly failed in intention-to treat analyses but may be valuable in delaying the need for surgery; further studies are warranted. Reports on outcomes following hyaluronic acid injections are contradictory. Conclusion No evidence exists on outcomes following non-operative management of FAI with concomitant Tönnis Grade 2 or more OA of the hip. Further studies are required and should explore the non-operative interventions that were employed for FAI and milder OA. There is strong evidence for a hip-specific physiotherapy program including activity modification and core strengthening exercises. Adjunct interventions such as corticosteroid injections and NSAID consumption may be valuable in delaying the need for surgery. Level of evidence Level IV. Supplementary Information The online version contains supplementary material available at 10.1007/s00167-022-07274-y.


Introduction
Femoroacetabular impingement (FAI) represents an abnormal hip morphology that may be a cause of development of hip osteoarthritis (OA) in view of the abnormal mechanical stresses being placed on the joint [21,37]. Usually, FAI is initially managed non-operatively [28], and this includes 1 3 activity modification/reduction, physiotherapy and analgesia. If symptoms persist, joint-preservation surgery can be considered [2,47]. However, outcomes of surgery for FAI with concomitant moderate-to advanced hip OA (Tönnis Grade ≥ 2) are equivocal, with some reports suggesting poor results and high rates of conversion to THA in the literature [13,15,22]. There is also concern that performing a jointpreservation procedure in this subgroup may lead to a rapid clinical decline that would require a THA sooner [52].
Previous randomized controlled trials (RCTs) (FASHIoN [24] and FAIT [45]) have successfully demonstrated good short-term clinical outcomes following non-operative management in patients with FAI and Tönnis Grade 0 and 1, even if it underperformed when compared to hip arthroscopy.
In a recent systematic review by our group [3,4], inconclusive and contradictory results were found for outcomes of hip arthroscopy for FAI with Tönnis Grade 2 or more hip OA. Whilst a significant amount of hip preservation surgeons consider Tönnis Grade 2 to be a contraindication for hip preservation surgery [17], some experts [10,11] reported favourable outcomes even in the context of moderate to advanced degeneration (Tönnis Grade 2 or more hip OA). These reports, however, did not include comparators with non-operative management.
In the context of a lack of consensus in the management of this specific patient cohort, the aim of the study was to investigate whether non-operative regimens alone may improve patient-reported outcomes (PROMs) and delay the need of subsequent surgery, either joint preservation or THA in patients with FAI and concomitant Tönnis Grade 2 or more hip OA.

Identification of studies
The review followed the PRISMA Extension for Scoping Reviews (PRISMA-ScR) [50] guidelines and used the recommendations outlined by Arksey and O'Malley [8].
The literature search of the PubMed database included articles from inception to 1st of December 2021 using combinations of keywords as shown in Supplementary Table 1. Studies reporting on PROMs and clinical or radiographic outcomes of non-operative management for patients with FAI were included for analysis. Reviews, case reports, surgical techniques, oral presentations and letters were excluded from the analysis. Two reviewers independently screened all the studies (AO and DK). In the case of a discrepancy, the senior author (VK) was consulted.
For the primary analysis, studies investigating FAI and concomitant hip OA Tönnis Grade ≥ 2 were considered eligible. A secondary aim was to identify available non-operative interventions and their outcomes for FAI irrespective of the degree of degeneration including all types of classification and stages for OA.
The following exclusion criteria were employed: studies including participants with active inflammatory disease, neurologic conditions, previous ipsilateral surgeries of the hip, osteonecrosis or concomitant hip dysplasia.

Data extraction
Data were extracted by two independent members of the review team. For each individual study of the secondary aim, information was recorded on an Excel Spreadsheet. The first author, name of the scientific journal, year of publication, and level of evidence according to the Oxford Centre for Evidence-Based Medicine 2011 Levels of Evidence [43] were extracted. Demographic and outcome data included the following: number of patients and hips, classification of osteoarthritis, Tönnis grading when available, indications, disease duration, type of intervention, age, body mass index (BMI), gender, length of follow-up, PROMs pre-and postoperative, radiological outcomes, rates of conversion to surgery or total hip arthroplasty, complications or adverse events and other interventions.
The extracted data were synthetized according to the level of evidence reported by the authors (Oxford Centre for Evidence-Based Medicine 2011 Levels of Evidence) [43] and type of non-operative intervention. At every step where there was a mismatch between reviewers during the screening and data extraction process, the senior author was contacted (VK) until agreement was reached.

Study selection
The database search yielded 962 studies. After removal of duplicates, a total of 548 articles were identified (Fig. 1). Five-hundred and seventeen studies were excluded at the title and abstract stage. Thirty-three full text papers were assessed for eligibility against the inclusion and exclusion criteria. Of these, not a single study met the inclusion and exclusion criteria. There was a single study that included patients with Tönnis Grade 2 OA, which however did not report outcomes for non-operative management separately for this cohort of patients [36].

Hyaluronic acid injections
One RCT [34] and two retrospective cohort series [1,44] utilised intra-articular hyaluronic acid (HAc) injections of the hip (Table 1) [1,34,44]. Abate et al. [1] also used activity restriction and allowed concomitant consumption of non-steroidal anti-inflammatory drugs (NSAIDs). The study reported slight improvement in PROMs (Table 1) at 12 months. Lee [34] compared hyaluronic acid injections (2 mL) with corticosteroid injections (20 mg triamcinolone acetate) and concluded after a mean follow-up of 3 months that no differences could be reported when separate injections were performed. In patients with crossover injection (both hyaluronic acid and corticosteroid), mean Hip disability Osteoarthritis Outcome Score (HOOS) increased from 51.7 at baseline to 71.0 at 3 months [34]. Ometti [44] registered significant improvement among all PROMs (visual analogue scale (VAS), Harris Hip Score (HHS), Lequesne) at 12 months after using a derivative of hyaluronic acid obtained by controlled chemical synthesis (2% partial hexadecylamide). Lee et al. [34] utilized fluoroscopic guidance as opposed to ultrasound guidance in the other two studies and were the only ones to report the dosages of injections.

Corticosteroid injections
One RCT [34], one prospective cohort study [46] and three retrospective case series [14,32,49] investigated the individual efficacy of corticosteroid injections either in a single cohort or compared to other interventions ( Table 2). Lee et al. also investigated the outcomes of HAc injections and crossover injections, for both HAc and corticosteroid [34]. As mentioned previously, PROMs improved only for patients that underwent crossover injections at 3 months. Another comparative study [46] reported no significant differences in outcomes between the three interventions that were assessed (activity, physiotherapy and hip arthroscopy) at 2 years of follow-up (Table 2).
Krych et al. [32] reported a 100% rate of conversion to hip arthroscopy in all 54 hips treated with corticosteroid injections at a mean time of 3 months post-intervention.   High rates of conversion have also been reported in the two remaining studies [14,49], averaging 50% (Table 2).

Hip bracing
Two studies with prospectively collected data (1 randomized controlled trial (RCT) [20] and 1 prospective cohort study) [42] evaluated the efficacy of a hip brace during a short period of follow-up (6 and 4 weeks, respectively). These have reported contradictory changes in PROMs (Table 3). Whilst Newcomb et al. [42] did not observe any significant changes in PROMs when compared to the standard care without the wearing of a hip brace, Eyles et al. [20] reported improvements for the HAGOS (The Copenhagen Hip and Groin Outcome Score) (p = 0.02) in the following categories: pain, symptoms, function and quality of life (QoL). However, these patients also received concomitant non-operative interventions with corticosteroid injections and oral NSAID consumption (Table 3).

Physiotherapy
Most studies that reported outcomes for non-operative interventions for FAI utilised different regimes of physiotherapy with or without other treatments ( Table 4). Out of the 16 studies in this subgroup [7,12,19,24,25,27,29,30,35,36,41,45,46,48,51,54], 5 randomized prospective studies [24,27,35,36,45] provided evidence for significant improvements post-intervention after hip-specific physiotherapy for patients with FAI, although underperforming when compared to hip arthroscopy (Table 4). There is also consistent reporting of significant positive impact on outcomes when employing core strengthening and postural exercises. A detailed breakdown of the exercise regimens and interventions is provided in Suppl. Table 2. The majority of studies also utilised activity modification with avoidance of positions with increased hip stress or restriction with appropriate patient education (Table 4). Seven studies [7,12,24,25,27,29,54] allowed concomitant interventions to be employed as a part of the non-operative regimen (corticosteroid injections and oral consumption of NSAIDs). Narveson et al. [41] emphasized the importance of therapeutic neuroscience education for patients with biopsychosocial issues. There was a significant variability among reports regarding the proportions of patients who ultimately underwent surgery, ranging from 8 [24] to 70% [35] ( Table 4). The average length of follow-up was heterogeneous and ranged from 2 [7] to 60 months [54]. Time to surgery was an equally heterogeneous parameter, with three studies reporting a mean time to surgery within 1 year after starting non-operative management (

Discussion
The main purpose of this scoping review was to identify studies that reported on the outcomes of non-operative management options for FAI with concomitant Tönnis Grade 2 or more hip OA. A significant number of studies were excluded whereby the primary aetiology was not described or in cases of concomitant hip dysplasia. Only Martin et al. [36] evaluated outcomes of non-operative management which included hips with FAI and Tönnis 2 OA. The authors, however, did not report outcomes separately for this patient cohort, but instead cumulatively with Tönnis 0 and 1 OA, precluding data extraction and analysis. As such, we conclude that there is currently no evidence on outcomes of non-operative interventions for FAI with Tönnis Grade 2 or more hip OA. Along with the findings of a recent systematic review that investigated the outcomes of hip arthroscopy [3] for the same patient cohort, our findings demonstrate the lack of reliable evidence that would guide clinicians in treating these patients. Additionally, there may have been studies that have reported the outcomes of non-operative management of such patients, but lacked disease stratification and may have used hip osteoarthritis as their indication [26]. The majority of studies describing non-operative management for hip OA do not describe the primary aetiology [9,23] that did not allow the inclusion of such studies. These studies may have provided valuable data, but due to increased selection bias and questionable primary aetiology (FAI/dysplasia or other causes of hip OA), we were not able to include them in the scoping review. The secondary aim of the scoping review was to explore all possible non-operative interventions for the population of interest. There is level I evidence with large-scale RCTs that have shown achievement of good PROMs after hip-specific physiotherapy and activity modification for FAI and mild OA [24]. Nevertheless, these studies had a relatively short follow-up time for up to 12 months. A recent study by Zogby et al. [54] has demonstrated that non-operative management can achieve mid-to long-term symptom improvement and significantly delay the need for subsequent surgery for patients with FAI and labral tears. There is also evidence pointing to the importance of implementing core strengthening and postural exercises in the physiotherapy regimen [25]. Casartelli et al. [12] and Aoyama et al. [7] compared two cohorts, one with standard physiotherapy focusing on hip stretching and hip muscle strengthening and the other that included core strengthening and found significant improvements in PROMs in the latter. Physiotherapy may also have a beneficial effect if patients have concomitant causes of extraarticular hip pain such as iliopsoas impingement [5] or iliotibial band tightness or snapping [6], by means of muscle stretching and improvement in muscle balance. The prevalence of concomitant extraarticular conditions (ischiofemoral [39], iliopsoas and subspine impingement [40]) in FAI and the role of physiotherapy in this cohort needs to be further investigated. An important adjunct to a successful physiotherapy programme is appropriate patient education. Apart from systematic in-person supervised training sessions, the authors [30,51] have emphasized the importance of patient education for home exercises and also techniques to increase patient compliance by reviewing patient progress using phone interviews or a training diary. Furthermore, Narveson et al. [41] have underlined the importance of additional therapeutic neuroscience education in patients that had biopsychosocial issues. A similar approach for hip arthroscopy, whereby surgeons were encouraged to screen patients for mental health disorders before surgery and offer them appropriate counselling, was described in a recent systematic review [16] due to associated poor surgical outcomes [33]. A therapeutic approach in this direction is currently lacking with only prognostic relationships being described.
Inconclusive and contradictory results were encountered when evaluating the efficacy of hip bracing. In a prospective study comparing hip bracing with physiotherapy as opposed to physiotherapy only, Newcomb et al. [42] found no differences in patient-reported outcomes (NRS pain, iHot-33 and HAGOS questionnaires) after 4 weeks of follow-up. Patients reported discomfort when wearing a brace (4.8/10 on NRS (numeric ratio scale). Although participants did not cease to wear the brace, there were one or more adverse events for each participant, the most common of which were knee irritation of the braced leg (12/17, 71%), slipping of the brace (8/17, 47%), brace-related discomfort during or after brace wear (6/17, 35%), and contralateral hip pain (4/17,24%) [42]. As opposed to these findings, Eyles and colleagues [20] achieved significantly better PROMs when compared with a physiotherapy-only cohort at 6 weeks of follow-up. However, the intervention also involved concomitant oral consumption of NSAIDs and corticosteroid injections. Although the authors reported a similar rate of usage between groups, the impact of hip bracing as an individual intervention is questionable. Even if considered effective, further research is required to show whether it provides any benefit beyond short-term follow-up.
Another subject of debate is the efficacy of injections with hyaluronic acid. Abate et al. reported slight improvement in PROMs from pre-to post-intervention at 12 months. The concomitant activity modification and consumption of oral NSAIDs raises concerns about the validity of the conclusion of the study. Lee et al. [34] compared hyaluronic acid injections with corticosteroid injections and concluded after a mean follow-up of 3 months that no differences could 1 3 be reported when injections were performed separately and individually. Ometti et al. [44] were the only to register significant improvement amongst all PROMs (visual analogue scale (VAS), Harris Hip Score (HHS), Lequesne) at 12 months after using a derivative of HAc. Recent research provides promising data regarding the potential of high molecular weight HAc injections as shown in a meta-analysis for hip osteoarthritis by Ebad Ali et al. [18] The authors concluded that short-term pain relief may be obtained in patients with hip degeneration.
A more valid modality for pain relief is injection with corticosteroids, as per a systematic review and meta-analysis by Zhao et al. [53] The authors compared all types of injection including HAc, corticosteroids, and platelet-rich plasma (PRP) for hip OA. When compared to our results, in the reports by Pennock [46] and Lee [34], corticosteroid injections achieved similar outcomes when compared to HAc injections, activity modification and hip arthroscopy. However, three studies [14,32,49] have mentioned a high proportion of patients undergoing surgery (from 50 to 100%) within 12 months after injection. These findings may support the hypothesis that the corticosteroid injection can provide short-term pain relief and possibly delay the need for THA, but do not ultimately alter the course of the disease process.
The limitations of the scoping review are directly linked to the limitations of the included studies. No evidence on outcomes for non-operative management for FAI with Tönnis Grade 2 or more OA was found. Studies that report on outcomes for these patients may have described the indication as hip OA, without describing the primary aetiology that may have been different from FAI (dysplasia, osteonecrosis, Perthes, posttraumatic or other) [23]. This did not allow inclusion of these studies for further assessment.
A lack of evidence of outcomes for non-operative management for patients suffering from FAI and concomitant Tönnis Grade 2 OA was found. Along with the findings of a recent systematic review [3] where inconclusive and contradictory evidence of outcomes of hip arthroscopy in this patient cohort were described, the scientific knowledge about the management of these patients remains insufficient and therefore lays the foundation for a randomised controlled trial. Although a THA may be a successful intervention in this specific cohort of patients with promising outcomes [31,38], further research is warranted in determining the efficacy in delaying the need for a THA.

Conclusion
There is a lack of evidence of outcomes following non-operative management of FAI with concomitant Tönnis Grade 2 or more OA of the hip. Further studies are needed and should explore the non-operative interventions that were employed for FAI and milder OA.
For mild OA, there is strong evidence for a hip-specific physiotherapy programme including activity modification and core strengthening exercises. Adjunct interventions such as corticosteroid injections and NSAID consumption may be valuable in delaying the need for surgery.