Operative Management of Avascular Necrosis of the Femoral Head in Skeletally Immature Patients: A Systematic Review

Purpose: Osteonecrosis of the femoral head (ONFH) is common in skeletally immature patients. The management of ONFH is controversial, with limited evidence and unpredictable results. This study systematically reviewed the current operative modalities and clinical outcomes of surgical management for ONFH in skeletally immature patients. Methods: The present study was conducted according to the PRISMA 2020 guidelines. PubMed, Google Scholar, Embase, and Web of Science databases were accessed in October 2021. All the published clinical studies reporting data concerning the surgical management of ONFH in skeletally immature patients were included. Results: This review included 122 patients (127 hips). 38.2% (46 of 122) were female. The mean age of the patients was 14.2 ± 2.3 years. The mean duration of the follow-up was 55.3 ± 19.6 months. The Harris Hip Score improved from 68.8 ± 11.9 at baseline to 90.5 ± 6.5 at last follow-up (p < 0.0001). Femoral head collapse and secondary hip degeneration were the most common complications. Conclusion: Several surgical techniques are available and effective for the management of ONFH in skeletally immature patients. This study evidenced high heterogeneity of the surgical procedures and eligibility criteria. Further high-quality investigations are required to establish proper indications and surgical modalities.


Introduction
Osteonecrosis of the femoral head (ONFH) is a disabling condition which leads to progressive pain, deformity, and early-onset osteoarthritis [1,2]. The growth plate acts as a barrier for the intraosseous blood vessels, and femoral head vascularisation is guaranteed by the lateral epiphyseal vessels [3][4][5]. Once bone maturity has been reached, the lateral epiphyseal vessels, the vessels of the metaphysis, epiphysis, and round ligament of the femur anastomose [3][4][5][6]. The aetiology of ONFH can be traumatic and atraumatic [7]. In traumatic ONFH, a fracture may cause an interruption of the blood supply to the femoral head [8]. Although proximal femur fractures in skeletally immature patients are rare, the risk of ONHF insurgence is high, and affects approximately 20% of such patients [9][10][11][12]. Several causes of atraumatic ONFH in the young population have been described; prolonged steroids consumption [7,13,14] and Legg-Calvé-Perthes disease [15,16] represent the most common etiological factors; sickle cell disease [17,18], lymphoblastic leukaemia [19,20], human immunodeficiency virus (HIV) infection [21,22] and Gaucher's disease [23] are other less common causes. The management of ONFH in skeletally immature patients Life 2022, 12, 179 2 of 12 is controversial, with limited evidence and unpredictable results [24]. Several operative strategies have been developed to regenerate or avoid further femoral head collapse, but no consensus has been reached. This study systematically reviewed the current available evidence on the surgical management of ONFH in skeletally immature patients, focusing on techniques, efficacy, and the safety profile.

Eligibility Criteria
All of the clinical studies evaluating the efficacy and safety of operative management of ONFH in skeletally immature patients were considered. According to the authors language capabilities, articles in English, German, and Italian were eligible. Only studies published in peer-reviewed journals were eligible. Only level I to IV of evidence articles, according to Oxford Centre of Evidence-Based Medicine [25], were considered. Reviews, opinions, letters, and editorials were not considered. Animals, biomechanics, computational, and cadaveric studies were not eligible. Only studies investigating a population with a mean age ≤18 years were eligible. Studies which involved less than five patients were excluded, as were those with a mean follow-up ≤24 months, or studies focusing on Legg-Calve-Perthes disease. Only studies which reported quantitative data under the outcomes of interest were considered.

Search Strategy
This systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses: the 2020 PRISMA statement [26]. The following algorithm was preliminary pointed out: In October 2021, the following databases were accessed: Pubmed, Web of Science, Google Scholar, EMBASE. No time constrains were used for the search. The following keywords were used: osteonecrosis, avascular, necrosis, collapsed, femur, femoral, head, fractures, ONFH, adolescent, paediatric, skeletally immature, surgery, operative, treatment, management, osteotomy. The keywords were combined using the Boolean operator AND/OR.

Selection and Data Collection
Two authors (F.M.; G.L.P.) independently performed the database search. All resulting titles were screened and the abstract of the articles which matched the topic were accessed. If the abstract matched the topic, the full-text article was accessed. A cross-reference of the bibliography of the full-text articles was also undertaken. Disagreements were debated and the final decision was made by a third author (N.M.).

Data Items
Two authors (F.M.; G.L.P.) independently performed data extraction. Generalities of the articles were retrieved: author and year, journal, duration of the follow-up, type of study. The following data were extracted: age, sex, aetiology, type of treatment, Harris Hip Score (HHS) [27], and complications.

Study Risk of Bias Assessment
The Coleman Methodology Score (CMS) was used to assess the quality of the methodological assessment [28]. Each article was assessed by a single reviewer (G.L.P.). The CMS is divided in two sections: A and B. Section A evaluated seven items: study size (number of Life 2022, 12, 179 3 of 12 patients), mean follow-up (from <12 month to >60 month), number of surgical approaches, type of study (randomized, prospective, or retrospective), and the description of diagnosis, surgical procedure, and rehabilitation. Section B evaluated the outcome criteria, method of assessing outcomes, and subject selection process. The final score ranges from 0 to 100 points, with values > 60.

Synthesis Methods
All statistical analyses were performed using the IBM SPSS Software version 25. For descriptive statistics, mean and standard deviation were calculated. To evaluate the improvement of the HHS from baseline to the last follow-up, the mean difference (MD) and T-test were performed. The confidence interval (CI) was set at 95%. Values of p < 0.05 were considered statistically significant.

Study Selection
The initial literature search resulted in 209 studies. Of them, 70 were excluded for being duplicates. Another 120 were not eligible, either for not matching the topic (N = 71), focusing on adults (N = 22), focusing on surgical management (N = 6), focusing on Legg-Calve-Perthes disease (N = 16), the type of study (N = 2), full text not accessible (N = 2), or uncertain results (N = 1). This left 19 articles for inclusion. An additional 11 studies were excluded as they did not report quantitative data under the outcomes of interest. This resulted in eight studies for analysis. The literature search results are shown in Figure 1.

Study Risk of Bias Assessment
The Coleman Methodology Score (CMS) was used to assess the quality of the methodological assessment [28]. Each article was assessed by a single reviewer (G.L.P.). The CMS is divided in two sections: A and B. Section A evaluated seven items: study size (number of patients), mean follow-up (from <12 month to >60 month), number of surgical approaches, type of study (randomized, prospective, or retrospective), and the description of diagnosis, surgical procedure, and rehabilitation. Section B evaluated the outcome criteria, method of assessing outcomes, and subject selection process. The final score ranges from 0 to 100 points, with values > 60.

Synthesis Methods
All statistical analyses were performed using the IBM SPSS Software version 25. For descriptive statistics, mean and standard deviation were calculated. To evaluate the improvement of the HHS from baseline to the last follow-up, the mean difference (MD) and T-test were performed. The confidence interval (CI) was set at 95%. Values of p < 0.05 were considered statistically significant.

Study Selection
The initial literature search resulted in 209 studies. Of them, 70 were excluded for being duplicates. Another 120 were not eligible, either for not matching the topic (N = 71), focusing on adults (N = 22), focusing on surgical management (N = 6), focusing on Legg-Calve-Perthes disease (N = 16), the type of study (N = 2), full text not accessible (N = 2), or uncertain results (N = 1). This left 19 articles for inclusion. An additional 11 studies were excluded as they did not report quantitative data under the outcomes of interest. This resulted in eight studies for analysis. The literature search results are shown in Figure 1.

Study Risk of Bias Assessment
The length of the follow-up was acceptable in most studies. Surgical technique, diagnosis, and rehabilitation protocols were generally well described. The study size and the retrospective design of most of the included studies represented the main limitations highlighted by the CMS. Outcome measures, assessment timing, and selection process were Life 2022, 12, 179 4 of 12 also clearly defined by most studies. Finally, the mean methodology score of 73.8 (range: 69 to 79) suggested an overall good quality of the methodological assessment (Table 1).

Author, Year Intervention Efficacy
Bartoníček et al., 2011 [2] Valgus-flexion intertrochanteric osteotomy (five hips) In all patients, the osteotomy healed within three months without complications. Limb shortening was fully corrected in all the patients. Radiographs and MRI scans after osteotomy proved resorption of the necrotic segment of the femoral head and its remodeling in all the patients.

Author, Year Intervention Complications
Bartoníček et al., 2011 [2] Valgus-flexion intertrochanteric osteotomy (five hips) One of the five patients was performed another osteotomy at one year after the surgery for the presence of a necrotic segment of the femoral head Bartoníček et al., 2012 [29] Valgus intertrochanteric osteotomy (six hips); Total hip replacement (five hips) VITO: one of six patients was reoperated five months after osteotomy for nonunion THR: no complications were observed Gatin et al., 2016 [30] Triple acetabular osteotomy (seven hips); Varus osteotomy (two hips); Combination of both (two hips) In one patient, femoral osteotomy was performed in addition to a triple acetabular osteotomy. Pneumonia occurred in a patient at two days postoperatively Ko et al., 1995 [31] Non-vascularized bone grafting via trapdoor procedure combined with containment osteotomy of femur and acetabulum (13 hips) Trapdoor procedure failed in two patients, one patient was subsequently treated by THA and the other one by femoral head allograft, at three and two years, respectively. Several patients had a mild to moderate limp but did not have pain.

Discussion
According to the main findings of the present study, surgical management is effective to improve symptoms, and decelerate or arrest the evolution of ONFH. Overall pain reduction, motion and gait improvement were reported by most patients. Surgical management aims to revitalize the femoral head, favour spherical remodeling, induce necrosis resorption, and improve hip motion.
The mean age of the population included for analysis was 14 years. At the approximate age of 14 years in girls and 16 years in boys, the growth plates of the femur start to fuse [44,45]. Hence, the population included was likely to still have open epiphyseal growth plates. Patients with open epiphyseal growth plates demonstrated higher bone marrow derived stem cell concentration compared to adults [46][47][48]. Moreover, the open physis promotes greater self-healing [49]. In this context, the use of bone marrow stimulating techniques should promote a greater regenerative potential in this population compared to adults. Further studies investigating the efficacy of bone marrow stimulating techniques of simple execution (e.g., isolated core decompression) in early ONFH should be considered. Multiple epiphyseal drilling augmented with autologous bone marrow implantation has been investigated in skeletally immature patients with ONFH secondary to sickle cell anaemia [34]. 29% (four of 14) of patients treated with multiple epiphyseal drilling augmented autologous bone marrow demonstrated significant improvements in symptoms and motion, and no further necrosis progression was evidenced in the half of the patients (seven of 14) [34]. Valgus and varus intertrochanteric osteotomy and triple acetabular osteotomy have been successfully performed in the early stages of ONFH to promote bone healing by discharging weight bearing over the necrotic area [2,29,30]. In patients with moderate to advanced ONFH, bone grafting may be performed [50][51][52]. Bone grafting replaces necrotic material, stimulates revascularisation, and, given its osteoinductive properties, promotes bone reconstruction [35,52,53]. The trapdoor procedure was effective in reducing the risk of femoral head collapse progression, improving clinical function in patients with moderate ONFH [32,54]. The trapdoor procedure aims to replace the necrotic area, promote revascularisation, and prevent collapse [31,54]. Ko et al. [31] investigated the potential of the trapdoor procedure in isolation and in combination with femoral and/or acetabular osteotomy on 13 young patients with advanced ONFH. The isolated procedure failed in two of three patients, while those operated with the combined procedure reported significant improvement in symptoms, activity, and range of motion [31]. These results indicated that the isolated procedure may be recommended for early-stage degeneration, and should be performed in combination in more advanced stages of ONFH. Zhang et al. [35] reported the outcome of 28 skeletally immature patients treated with a free vascularised fibular graft. Regardless of the extent of ONFH, all patients experienced an excellent outcome, with an improvement in HHS from 60.4 to 94.2 points on average [35]. In patients with advanced ONFH, THA improved joint function and symptoms, restored leg length difference, and led to the greatest improvement of the HHS [35]. However, despite its excellent results, THA should be performed only as last resort, as long-term survival of the implants is not ensured [55][56][57].
Current evidence on ONFH management refers almost exclusively to the adult population or mixed cohort populations. Rotational osteotomies, core decompression in isolation or in combination with vascularised bone grafts, bone morphogenetic proteins, bone marrow-derived cells, mesenchymal stem cells, or combinations of the above, are the most commonly reported surgical procedures performed for early to moderate ONFH [58][59][60][61][62][63][64][65][66]. Total hip replacement is recommended for end stage ONFH or in cases of failure of the index procedures [67][68][69]. Core decompression augmented with autologous bone marrowderived transplantation or combined with autologous bone graft are effective in reducing symptoms of ONFH and the need for arthroplasty in the adult population [63,64]. The regenerative potential of autologous mesenchymal stem cells have been poorly investigated in the younger population, and further investigations are required.
This study certainly has limitations. The reduced number of included studies and consequently analysed procedures do not allow for the provision of accurate evidence-based recommendations. The cohort of included patients is not homogeneous in terms of pathogenesis of ONFH. Proximal femur fractures [2,29,[31][32][33]35], sickle cell anemia [30,31,34], idiopathic causes [31], infection [33], steroid use [31], lymphocytic leukemia [31], and post-adrenal tumor excision [31] are the most common causes of ONFH reported. With regard to proximal femur fractures, transepiphyseal, transcervical, cervicotrochanteric, and intertrochanteric fractures were reported. Furthermore, most authors used the HHS for patient evaluation. However, this score has not been validated for skeletally immature patients. When considering paediatric populations, child-friendly patient reported outcomes measures (PROMs) should be used. Most of the included studies were retrospective, with the lack of a control group and limited information. Given the lack of quantitative data, bilaterality of ONFH it was not possible to investigate separately. Finally, articles concern-Life 2022, 12, 179 9 of 12 ing Perthes disease have been excluded, as the management of these patients has been extensively analysed in the literature [70][71][72][73][74][75][76][77][78][79][80][81][82], and evidence on non-Perthes related ONFH is limited. Two studies which have been included for analysis reported data on patients with Perthes disease [31,33]; however, since the authors reported the results separately, data from patients with Perthes disease were not considered. Ko et al. [31] included for analysis 12 patients (13 hips). Of them, nine patients (nine hips) were affected by Perthes disease, and three patients (four hips) by idiopathic ONFH. In the present study, only the patients with idiopathic ONFH were considered for analysis. Current evidence on the surgical management of ONFH in skeletally immature patients is hence very limited. Therefore, larger studies should be undertaken to establish clearer indications and protocols in the paediatric population.

Conclusions
Several surgical techniques are available and effective for the management of ONFH in skeletally immature patients. This study evidenced high heterogeneity of the surgical procedures and eligibility criteria. Further high-quality investigations are required to establish proper indications and surgical modalities.

Funding:
The authors received no financial support for the research, authorship, and/or publication of this article.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The datasets generated during and/or analysed during the current study are available throughout the manuscript.