Osteonecrosis of the Femoral Head: Evolution of Contemporary Management Strategies

Abstract

Purpose of Review

Osteonecrosis of the femoral head (ONFH) is a progressive condition disproportionately affecting young adults and is a leading indication for total hip arthroplasty in young adults. This review aims to provide a contemporary overview of ONFH with emphasis on treatment strategies aimed to prevent femoral head collapse and disease progression as supported by recent clinical research from the past five years.

Recent Findings

The 2019 revised ARCO classification has emerged as the most clinically applicable staging system, guiding treatment selection based on lesion size, location, and collapse status. Core decompression remains the surgical mainstay for pre-collapse disease. Recent evidence supports the addition of hip arthroscopy to improve clinical outcomes and increase native hip survival rates. Augmentation strategies including bone marrow aspirate concentrate (BMAC), mesenchymal stem cells (MSCs), bone grafting, and hip arthroscopy have demonstrated the most promising results when combined with core decompression in early-stage disease. Pharmacologic agents, including bisphosphonates and denosumab, have shown meaningful benefit in slowing disease progression but remain adjunctive. Total hip arthroplasty continues to demonstrate excellent long-term outcomes for post-collapse stages, even in younger patient populations.

Summary

Management to prevent collapse and disease progression of ONFH remains a challenge. Joint preservation through early surgical intervention with augmentation strategies, such as addition of hip arthroscopy, bone grafting, and biologic regenerative biologic therapies represent promising advancements in ONFH management. However, the existing literature remains limited by heterogeneous study designs, small sample sizes, and inconsistent outcome reporting.

Introduction

Osteonecrosis of the femoral head (ONFH), previously termed avascular necrosis, is a progressive pathologic process characterized by compromised blood supply to the femoral head, resulting in osteocyte death, structural collapse, and eventual degenerative arthritis. ONFH disproportionately affects young and middle-aged adults and remains a leading indication for total hip arthroplasty (THA) in patients under 50 years of age [1, 2].

The epidemiology of ONFH is multifactorial, with both traumatic and atraumatic etiologies. Traumatic causes, including femoral neck fractures and hip dislocations, are less common and directly disrupt the vascular supply to the femoral head. Atraumatic ONFH is more prevalent and can be associated with corticosteroid use, chronic alcohol consumption, tobacco use, hemoglobinopathies such as sickle cell disease, autoimmune disorders, and metabolic syndromes [3–5]. Further, there is recent evidence suggesting that gastrointestinal dysbiosis may predispose to ONFH due to systemic effects on immunology and metabolic processes [6]. Despite these associations, it is estimated that 20 to 40% of cases remain idiopathic, suggesting a complex interplay between genetic predisposition and environmental risk factors [2].

While the pathophysiology of ONFH is not fully defined, it is widely accepted to involve vascular compromise leading to ischemia and impaired bone remodeling. Proposed mechanisms include intravascular coagulation, fat embolism, increased intraosseous pressure, and endothelial dysfunction [4, 5, 7]. These processes culminate in subchondral bone weakening and eventual collapse.

Traditional management of ONFH is dependent upon stage of disease. Core decompression remains the cornerstone of early-stage ONFH, while THA is a reliable surgical option for late-stage disease. Over the past decade, there has been substantial interest in biologic augmentation and minimally invasive techniques aimed at improving femoral head preservation in the management of early-stage ONFH [8].

This review aims to provide a contemporary overview of ONFH with emphasis on treatment strategies aimed to prevent femoral head collapse and disease progression as supported by recent clinical research from the past five years.

Diagnosis and Classification

Diagnosis of ONFH can be made by plain radiographs, CT scan, or MRI. The most accurate diagnostic test continues to be MRI, with sensitivities and specificities greater than 90% [1, 3, 4]. The classic radiographic findings include a radiolucent lesion surrounded by a serpentine sclerotic rim in the epiphysis or a thin linear subchondral lucency, commonly referred to as the crescent sign. The typical MRI pattern demonstrates a rim of low signal intensity on all sequences with an inner rim of high intensity found on fluid-sensitive sequences, which is known as a double-line sign [9, 10]. The kerboul angle is another parameter used to predict the volume of necrotic lesion in ONFH. It was initially adapted to plain radiographs, but was later modified to be measured on MRI for improved accuracy [11, 12]. The modified kerboul angle measures the combination of the largest area of necrosis identified on mid-coronal and mid-sagittal slices on MRI imaging. Ha et al. created a grading system based on modified kerboul angle that was designed to help predict progression of femoral head collapse [13]. Grade I is a combined angle < 200 degrees, grade II between 200 and 249 degrees, grade III between 250 and 299 degrees, and grade IV > 300 degrees combined angles. In this study, Ha et al. demonstrated all 25 hips with a combined angle > 240 degrees led to femoral head collapse.

Accurate staging of ONFH is critical for guiding treatment decisions and predicting outcomes. The Ficat and Arlet classification remains one of the earliest and most widely recognized systems, relying primarily on radiographic and clinical findings (Table 1) [14].

Table 1.

Modified Ficat & Arlet classification system

Stage	Findings
0	No clinical or radiographic findings; cellular changes present on biopsy only.
I	Plain radiographs are normal; MRI demonstrates bone marrow edema. Groin/hip pain may be present clinically.
II	Sclerosis and cystic changes without subchondral lucency or collapse.
III	Defined by subchondral regional collapse with outer sequestrum (“crescent sign”).
IV	Complete collapse of the femoral head with secondary degenerative arthritis.

Findings are based on plain radiography unless otherwise specified. Stage 0 is a pre-symptomatic, histological stage not detectable by imaging

The Steinberg classification aimed to overcome the limitations of the Ficat and Arlet classification system by allowing more accurate identification of disease progression (Table 2) [15]. Benefits of the Steinberg system includes increased reliability by relying on radiographic parameters, rather than clinical symptoms. It also improved quantification of femoral head involvement by including three grades (mild, moderate, and severe) of severity. Given the correlation of femoral head involvement and risk of femoral head collapse, prognostication was improved with the Steinberg system. However, there are limitations with the Steinberg classification, as there is limited ability to guide treatment recommendations.

Table 2.

Steinberg classification system

Stage	Findings
0	Normal radiographs, bone scan, and MRI.
I	Normal radiographs; bone scan and/or MRI abnormal, consistent with osteonecrosis. A: Femoral head involvement < 15% B: 15%–30% femoral head involvement C: Femoral head involvement > 30%
II	Cystic and sclerotic changes on radiographs; no subchondral collapse present. A: Femoral head involvement < 15% B: 15%–30% femoral head involvement C: Femoral head involvement > 30%
III	Subchondral collapse produces “crescent sign”; femoral head contour remains spherical. A: < 15% of articular surface with underlying crescent B: Crescent present beneath 15%–30% of articular surface C: > 30% of articular surface with underlying crescent sign
IV	Flattening of the femoral head now present. A: < 15% surface collapsed and depression < 2 mm B: 15%–30% collapsed or 2–4 mm depression C: > 30% collapsed or > 4 mm depression
V	Joint space narrowing with or without acetabular involvement. A/B/C: Average of femoral head involvement (as in Stage IV) plus estimated acetabular involvement
VI	Advanced degenerative changes of the hip are present.

Subclassifications A, B, and C indicate mild, moderate, and severe disease, respectively; subclassification criteria for Stages I and II are identical. Subclassification of Stages I and II are facilitated with cross-sectional imaging. Progression to Stage IV is considered the threshold beyond which joint-preserving procedures are unlikely to be efficacious

Given the limitations of existing classifications, the Association Research Circulation Osseous (ARCO) was initially developed in 1992 with the goal of providing guidance in the treatment of ONFH [4]. In 2019, the ARCO classification underwent its latest revisions (Table 3). This iteration solidified 4 stages paralleling the Ficat stages while preserving integration of MRI findings and lesion location [16]. ARCO staging additionally includes lesion size and location (medial, central, lateral), which has significant prognostic implications. It has been demonstrated that lesions with the worst prognosis are those with a necrotic lesion volume > 30% of the femoral head and are located along the lateral pillar of the weight bearing surface of the femoral head [17]. Pre-collapse lesions (Stages I–II) are typically managed with joint-preserving strategies such as core decompression, while post-collapse disease (Stages III–IV) often necessitates arthroplasty.

Table 3.

Revised 2019ARCO staging criteria

Stage	Imaging Characteristics
I	No identifiable radiographic changes. T1-weighted MRI sequences show a hypointense band formed around the region of necrosis.
II	Focal areas of osteoporosis or cystic changes with adjacent sclerosis present in the femoral head.
III-a	Fracture or subchondral collapse of the necrotic region ≤ 2 mm.
III-b	Depression or flattening of the femoral head > 2 mm.
IV	Hip osteoarthritis as evidenced by joint space narrowing, fragmentation of the femoral head, and acetabular degeneration.

Findings are based on plain radiography unless otherwise specified. Stage I is detectable on MRI only

Nonoperative Management

Nonoperative management is generally reserved for early-stage disease, patients who are not surgical candidates, or those who wish to delay an appropriately indicated arthroplasty procedure. Standard nonoperative strategies include activity modification, protected weight-bearing, non-steroidal anti-inflammatory drugs (NSAIDs), and physical therapy. While nonoperative treatments may alleviate symptoms, they have a limited impact on disease progression, as 79% of conservatively managed early-stage ONFH hips have been shown to develop femoral head collapse at minimum 24-month follow-up [18].

Pharmacologic Therapies

Pharmacologic treatments have been evaluated for the early stages of ONFH with the goal of identifying non-surgical alternatives that either slow or reverse the progression of femoral head collapse and osteonecrosis [19]. Bisphosphonates and denosumab carry the most robust support in the literature for the prevention of femoral head collapse progression in early-stage ONFH.

Bisphosphonates have been extensively studied for their potential to inhibit osteoclast-mediated bone resorption (Table 4) [19]. Bisphosphonates act by inhibiting osteoclastic activity and reducing bone turnover, thus preventing breakdown of bone and further femoral head collapse. Bisphosphonate medications offer the most significant impact on slowing disease progression in early-stage ONFH. Lai et al. demonstrated that 25 weeks of weekly oral alendronate significantly reduced the rate of femoral head collapse (6.9% versus 76%, p < 0.001) and conversion to THA (3.4% versus 64%, p < 0.001) when compared to control groups receiving a placebo at 24-month follow-up [20]. A more recent retrospective review by Agarwala et al. compared disease progression in patients receiving alendronate biweekly in isolation to patients receiving bisphosphonate combination therapy with alendronate biweekly and zoledronate yearly at a minimum of 5 year follow up [21]. Authors demonstrated that both alendronate alone and bisphosphonate combination therapy prevented surgical intervention at a high rate (74.5% and 88.9%, respectively). Both groups were better at preventing disease progression at an early stage (ARCO I and II), but the failure rate increased to approximately 30% in both groups at ARCO stage III.

Table 4.

Adjunctive pharmacologic treatment

Medication	Mechanism of Action	Benefit
Bisphosphonates	Bind hydroxyapatite in bone to inhibit osteoclasts.	Can prevent progression of early disease stages.
Denosumab	RANK-ligand inhibitor which impedes osteoclast activity.	Reduces area of necrosis and prevents femoral head collapse.
Calcium Dobesilate	Vasoactive agent that reduces intravascular coagulation through multiple pathways.	Reduces bone marrow edema improving symptoms and delaying disease progression in early stages.
Statins, anticoagulants, vasodilators	Theoretically combat the vascular dysfunction involved in idiopathic ONFH.	Remains investigational without definitive evidence of efficacy.

Abbreviations: RANK receptor activator of nuclear factor kappa-B

Denosumab, a RANK-Ligand inhibitor, has emerged as a potential pharmacologic alternative for management of ONFH, as preliminary studies suggest improved bone remodeling and reduced progression of femoral head collapse (Table 4) [22–23]. Moon et al. evaluated 146 hips that received denosumab 60 mg subcutaneously every 6 months for one year and 126 hips that only received pain medications in the non-operative management of ARCO stage I and II ONFH with one year follow-up [22]. Authors identified a significant decrease in the rate of femoral head collapse when receiving denosumab compared to control (24.7% versus 38.1%, p = 0.012), but no difference in conversion to THA (15.6% versus 23%, p = 0.086). The denosumab group also demonstrated a significant reduction in the necrotic area observed on MRI compared to the control. The median relative change in necrotic lesion volume was a decrease of 15.5% in the denosumab group, in contrast with a 2.5% increase in the control group. Further, a retrospective study by Liu et al. evaluated a similar treatment dose of denosumab compared to a control group, but with a 2-year follow-up. Authors demonstrated significantly lower rates of progression of femoral head collapse in the denosumab group compared to the control group (42.2% versus 54.1%, p = 0.024) [23].

Calcium dobesilate has been shown to reduce platelet aggregation, decrease capillary permeability, and blood hyperviscosity while enhancing the endothelium-dependent vascular relaxation (Table 4). An early case study of 6 patients by Ransohoff et al. showed symptomatic improvement and prevention of disease progression on MRI with the medication [24]. Overall, with limited small-sample human studies completed, further studies are needed before any conclusions can be drawn.

Additional pharmacologic agents currently being investigated include statins, anticoagulants, and vasodilators (Table 4). While these therapies target specific pathogenic mechanisms, current evidence does not support their routine use as standalone treatments [19]. There is no evidence that the available pharmaceutical options effectively reverse findings of osteonecrosis or slow progression.

Surgical Management

While a role for non-operative and pharmacologic management for early-stage ONFH (ARCO I & II) exists, surgical management with core decompression remains the gold standard treatment. Other surgical options that have been studied include hip arthroscopy and osteotomies. Late-stage ONFH (ARCO III & IV) is generally accepted to be best treated with total hip arthroplasty [1–4].

Core Decompression

Core decompression (CD) remains the mainstay of treatment for pre-collapse ONFH (ARCO Stage I and II). Core decompression was first described by Ficat and Arlet in 1964 and has evolved over the years. Originally described as a single large-diameter drilling technique, modern CD has evolved to include multiple small-diameter drillings and percutaneous approaches aimed at minimizing morbidity while maximizing decompression [25].

The original technique of single large-diameter drilling includes passing a guide wire just superior to the lesser trochanter, centered in the femoral neck, and directed at the area of necrosis, which is often anterolateral [25]. An 8–10 mm cannulated trephine drill bit is then passed to core out the necrotic lesion. In early techniques, the core of bone removed was discarded. More recent single large-diameter drilling techniques include backfilling the cored-out segment with bone grafts or adjuvants [26]. Previous studies have shown increased risk of major complications with single large-diameter drilling techniques, most notably fracture and progression of disease to complete femoral head collapse, while demonstrating increased risk of osteonecrosis progression with larger areas of necrosis [27–29].

A multiple small-diameter drilling technique for core decompression was introduced in 2003 [30]. The technique uses one smaller drill bit, often 3.2 mm, to drill multiple different paths from a single (or a few) entry point into the necrotic lesion. Generally, more drill paths are used as the necrotic lesion increases in size. Multiple small diameter drilling reduces the rate of major complications compared to single diameter drilling alone, without grafting or adjuvants, but has not been shown to significantly affect the rate of progression of femoral head collapse [25].

More recentresearch has evaluated the role of adding hip arthroscopy, various bone grafting options, and adjuvant orthobiologic options to traditional drilling techniques in management of pre-collapse ONFH [25, 31]. Bone grafting options are classically divided into vascularized and non-vascularized grafts, as well as autograft and allograft options. A variety of adjuvant cell-based therapies have been developed, including bone marrow aspirate concentrate (BMAC), mesenchymal stem cells (MSCs), bone morphogenic proteins (BMP), and platelet-rich plasma (PRP).

Hip Arthroscopy

Hip arthroscopy (HA) has gained attention as an adjunct to core decompression, as it allows for direct visualization of intra-articular pathology and facilitates treatment of concomitant conditions such as labral tears, cartilage injury, and abnormal bony morphology. Additionally, HA can assist in the accurate placement of decompression tunnels. Results of current studies demonstrate promising results in the adjunctive use of HA with core decompression (Table 5).

Table 5.

Summary of key studies on the role of hip arthroscopy

Authors	Study Design	Methods	Participants	Summary of Key Findings
Yang et al., (2024) [31]	Retrospective comparative study	ARCO Stage II–IIIA; arthroscopic “light bulb” CD vs. traditional CD; min. 2-yr follow-up	39 pts (18 arthroscopic; 21 traditional)	• Arthroscopic group: lower VAS at 3 days post-op (2.0 vs. 2.8, p<0.05); higher HHS at last follow-up (80.1 vs. 75.1, p<0.05) • Lower THA conversion rate (11.1% vs. 28.6%); most common intraoperative findings: cartilage wear (83%) and synovitis (67%) • No major complications in either group
Ji et al., (2023) [30]	Prospective single-arm case series	ARCO Stages I–III; arthroscopic synovectomy + multi-hole CD; mean follow-up 10.7 yrs	103 patients / 153 hips	• Overall femoral head survival 51.6%; survival by stage: I - 79%, IIa - 72%, IIb - 52%, IIc - 32%, III 10% • Significant improvement in modified HHS and VAS from baseline (p<0.05); synovitis identified in 83% of hips • Multiple small-diameter holes preserved femoral head biomechanical integrity; no fractures or infections
Zhao et al., (2023) [32]	Retrospective comparative cohort	ARCO Stages I–II; arthroscopic intracapsular decompression + multi-channel CD (Group A) vs. CD alone (Group B); min. 2-yr follow-up	101 pts / 139 hips (Group A: 59; Group B: 80)	• Group A: significantly lower VAS at all time points through 1 year; higher HHS at 3 and 12 months (85.1 vs. 77.2; 81.9 vs. 76.4) • IAP significantly reduced post-op in Group A only (25.5 → 7.3 mmHg; p<0.001); no change in Group B • 2-year femoral head non-collapse rate: 89.8% (Group A) vs. 72.5% (Group B) (p=0.035); no major complications
Domb et al., (2025) [29]	Systematic review (PRISMA)	6 studies (4 Level III, 2 Level IV); CD with concomitant hip arthroscopy vs. without; min. 2-yr follow-up	632 hips (408 CD + arthroscopy; 224 CD alone)	• CD + arthroscopy yielded significantly greater postoperative HHS (71.41–93.33) across all comparative studies • 3 of 4 comparative studies: lower THA conversion rates and femoral head collapse with arthroscopy (10.2–23.8% vs. 25.5–28.6%) • THA-free survivorship 51.6–89.8% with arthroscopy; benefits attributed to intra-articular pathology treatment and IAP reduction

Abbreviations:IAP intra-articular pressure, yr year(s)

A meta-analysis by Domb et al. including six articles that evaluated 632 hips demonstrated improved patient-reported outcome measures (PROMs) and increased survivability, with lower evidence of femoral head collapse progression (10.2–23.8% versus 25.5–28.6%) and conversion to THA in combination HA and CD, compared to CD alone [32]. Longer term benefits have been demonstrated by Ji et al. as there was greater than 72% survivorship at 10-year follow-up in ARCO stage IIB ONFH treated with HA and CD [33].

The benefit of adding HA to CD is likelyrelated to the ability to beto address concomitant intra-articular pathology simultaneously. It is unclear whether there are direct benefits of HA in improving bone healing in ONFH, though some hypothesize that direct visualization during arthroscopy allows for more accurate drill tunnel placement at the site of osteonecrosis, thereby improving outcomes with more accurate decompression. Techniques such as the “light bulb” technique, have been developed to directly visualize decompression of the necrotic lesion. The “light bulb” technique initially approached the femoral head-neck junction from an open direct anterior approach, making an anterior 1 cm cortical window in the femoral head-neck junction to perform the decompression. It has since been adapted to a technique utilizing hip arthroscopy to access the same location of decompression in a less invasive manner [34]. Yang et al. were able to demonstrate that, compared to traditional CD alone, the “light bulb” technique of arthroscopic decompression at the femoral head-neck junction significantly improves VAS and Harris Hip Scores (HHS) [34]. Authors also demonstrated that significantly fewer patients in the “light bulb” group had radiographic progression of osteonecrosis requiring total hip arthroplasty compared to CD with bone grafting at two-year follow-up (11.1% vs. 28.6%). While these studies demonstrate decreased progression of osteonecrosis with addition of HA, studies have not demonstrated improved healing rates. Further, arguments against concomitant HA include increased operative time and increased cost [35].

Overall, larger studies are necessary to evaluate the true impact of HA on the management of early-stage ONFH. However, early smaller studies seem to suggest that, when combined with core decompression, clinical outcomes and survivability are improved in the treatment of early-stage ONFH with the addition of HA.

Bone Grafting

Bone grafting for ONFH is utilized to provide structural support and enhancebiologic healing and can be divided into vascularized versus non-vascularized and allograft versus autograft options. Bone graft options utilized in ONFH include iliac crest grafting, impaction grafting using cored out bone from decompression, non-vascularized fibular grafts, and free vascularized fibular grafts (FVFG) [36].

A retrospective study by Jie et al. evaluating 153 hips compared two groups of fibular grafts, autologous and allogenic, in combination with CD for early-stage ONFH (Table 6) [37]. The autologous and allogenic fibular allografts demonstrated increased HHS of 15.3 and 16.4 and Visual Analog Scores (VAS) scores of 4.0 and 3.9, respectively without a significant difference between either fibular graft. Both groups demonstrated an approximate 85% survival rate at 15-year follow-up, indicating high survivability when using either autogenous or allogenic fibular grafts in combination with CD.

Table 6.

Summary of key studies on bone grafting

Authors	Study Design	Methods	Participants	Summary of Key Findings
Wan et al. (2022) [33]	RCT	Randomized 4-arm trial comparing FFG, FVFG, ABG, and β-TCPG, each combined with minimally invasive CD. Outcomes: HHS, VAS, and radiographic ARCO staging	182 patients; 192 hips Follow-up: 42–48 months	• All four graft types produced significant improvements in HHS and VAS from baseline (all p<0.01); no significant difference between groups at final follow-up. • β-TCPG demonstrated significantly shorter operative time and lower blood loss compared to other groups. • Femoral head collapse at last follow-up was low and comparable across groups without statistically significant differences (3 FFG, 2 FVFG, 2 ABG, 3 β-TCPG) • β-TCP bioceramic graft identified as a viable alternative to autologous grafts for early-stage ONFH
Jie et al. (2021) [34]	Retrospective cohort	Comparison of CD combined with non-vascularized autologous fibular graft vs. allogeneic fibular graft; ARCO stages IIa–IIIc. Outcomes: HHS, VAS, FJS. Survival analysis via Kaplan–Meier; Endpoint: THA conversion.	117 pts; 153 hips Autologous group: 34 pts, 50 hips; Allogenic group: 83 pts, 103 hips; mean f/u 9.3 yr	• Both groups showed significant postoperative improvement in HHS and VAS (p<0.05); no significant intergroup differences in HHS, VAS, FJS, or patient satisfaction at final follow-up. • 15-year survival rates were comparable: autologous 84.1% vs. allogeneic 86% (p=0.355). • Autologous grafting associated with longer operative time, greater blood loss, higher material cost, and donor-site morbidity (peroneal nerve injury 6%; donor numbness 10%). Allogeneic grafting carried a 30% rate of transient postoperative fever from immune rejection. • Patient age >37 years was the only significant independent risk factor for THA conversion (HR 6.93, 95% CI 2.00–24.07). • Mean time to THA conversion was longer in the autologous group (11.2 yr) vs. allogeneic (4.1 yr), possibly reflecting early immune-mediated disruption of the repair environment.
Kuroda et al. (2021) [35]	Retrospective observational	Evaluation of percutaneous autologous IBG modification following. Outcomes: HHS, UCLA activity scale, Kaplan–Meier survival (Endpoint: THA conversion), and radiographic staging by JIC classification.	20 patients; 20 hips Mean f/u: 36 months 19 patients post collapse	• Technique was minimally invasive with mean operative time <1 h and minimal blood loss; no intraoperative complications. • 35% of patients (7/20) required THA conversion at a mean of 17 months postoperatively; 3-year Kaplan–Meier survival rate was 65%. • Radiographic progression observed in 60% of patients, though 6 patients showed radiological worsening without corresponding clinical deterioration. • Among patients not converting to THA, HHS and UCLA scores improved significantly at 3 years (HHS: 57.6 → 76.5, p=0.005; UCLA: 3.7 → 5.2, p=0.014). • Short-term outcomes unsatisfactory for advanced ONFH; strict patient selection and adjunctive strategies (additional autologous bone, growth factors, modified rehabilitation) are needed.
Hu et al. (2023) [36]	Systematic review / Network meta-analysis (NMA)	Analysis of RCTs evaluating 11 surgical interventions for ONFH. Outcomes: ONFH progression rate, THA conversion, and HHS improvement.	18 RCTs 1,107 hips total Follow-up: 12–36 months	• Compared with CD alone, ABG, FFG, VBG, ABG+BMAC, and BMG+VBG all significantly delayed ONFH progression. • ABG+BMAC ranked highest for preventing ONFH progression (OR 0.019; 95% CI 0.0012–0.25; SUCRA=0.926), suggesting that combining mechanical support with biological regenerative capacity is most effective. • All treatments incorporating bone grafting were effective in preventing ONFH progression, underscoring the fundamental importance of structural support at the necrotic zone regardless of graft type.

Abbreviations: IBG impaction bone graft, JIC Japanese Investigation Committee, mo months, SUCRA surface under the cumulative ranking, UCLA University of California, Los Angeles, β-TCP beta-tricalcium phosphate, β-TCPG β-TCP bioceramic graft

A prospective randomized controlled trial by Wan et al. evaluated 192 hips that underwent core decompression and were then divided into four groups receiving either autogenous fibular graft, free vascularized fibular graft, autologous iliac bone graft, or beta-tricalcium ceramic phosphate graft (β-TCPG) [36]. The authors demonstrated no statistical difference between the treatment groups in rate of progression of femoral head collapse or PROMs, though all four groups did demonstrate improved preoperative to postoperative PROMs. The Authors also demonstrate less blood loss and marginally faster operating times with B-TCPG compared to other treatment groups.

Kuroda et al. evaluated 20 hips retrospectively that received CD with autogenous impaction grafting using the cored-out bone from the CD procedure [38]. The authors demonstrated high rates of progression of femoral head collapse (60%, 12/20 hips) and conversion to THA (35%, 7/20 hips) in their cohort over a mean 3-year follow-up, suggesting impaction grafting alone is not sufficient to treat early-stage ONFH.

In comparative studies, bone grafting has demonstrated improved clinical outcomes when compared to CD alone [25, 39]. However, there does not appear to be a superior bone grafting option in the existing literature. It is reasonable to consider using a bone grafting option to supplement core decompression in the management of pre-collapse ONFH, though factors such as cost should be taken into consideration (Table 6).

Osteotomy

Osteotomy techniques aim to redistribute necrotic regions of the femoral head out of the weight-bearing zone. Common procedures include curved intertrochanteric varus osteotomy, trochanteric rotational osteotomy, and half wedge osteotomy [40]. While less commonly performed in Western populations, osteotomy remains a viable option in young patients with localized lesions and has shown 10-year survival rates upward of 70% [41]. The most significant complication and cause for revision surgery following osteotomy is the progressive collapse of the femoral head [42]. These risks are important to weigh against the continued improvement in THA survival rates among younger patients [43–44].

Arthroplasty

Total hip arthroplasty (THA) remains the definitive treatment for post-collapse ONFH (ARCO stage III and IV). Studies have demonstrated excellent pain relief and functional outcomes following THA for ONFH [44–45]. Favre et al. evaluated 290 THAs in patients younger than 50 that included propensity-matched cohorts of ONFH and osteoarthritis (OA). The authors demonstrated excellent 20-year survivability for both OA (79.8%) and ON (83.4%) [44]. While there has been concern for implant failure long term, advances in implant technology and bearing surfaces may lead to improved survivorship, making THA a reliable option in late-stage disease, even in younger populations [46]. Extensive literature exists and continues to support the utility of THA as the end-of-line treatment in osteonecrosis of the femoral head.

Adjuvant Therapies

Adjunctive therapies represent one of the most rapidly evolving areas in ONFH management. Adjuvant therapies are added tocore decompression with the goal of improving biology at the site of necrosis by promoting bone healing, as opposed to simply preventing further femoral head collapse. A variety of different adjuvant therapies have gained attention in recent years, including bone marrow aspirate concentrate (BMAC), mesenchymal stem cells (MSCs), platelet-rich plasma (PRP), bone morphogenic proteins (BMP), hyperbaric oxygen therapy (HBOT), and intraosseous medications including bisphosphonates (Table 7) [47].

Table 7.

Summary of key studies on adjuvant treatments combined with core decompression

Authors	Study Design	Methods	Participants	Summary of Key Findings
Zhang et al. (2020) [49]	Systematic Review & Meta-Analysis	Analyzed 16 studies (7 RCTs; 9 comparative) comparing CD alone vs. CD + BMSCs for early-stage ONFH. Outcomes: VAS pain score, HHS, stage progression rate, and THA conversion rate.	16 studies; 583 hips (CD + BMSCs) vs. 468 hips (CD alone)	• CD + BMSCs produced significantly lower VAS pain scores at 24 months (MD = −10.88; p = 0.003) and higher HHS (MD = 5.59; p = 0.01) vs. CD alone. • Stage progression: 27.0% (CD + BMSCs) vs. 51.4% (CD alone); RR = 0.51 (p = 0.0002). • THA conversion: 22.5% (CD + BMSCs) vs. 42.3% (control); p = 0.001. • Despite heterogeneous evidence, BMSCs combined with CD demonstrate meaningfully better clinical and radiographic outcomes than CD alone.
Aggarwal et al. (2021) [54]	Prospective RCT	Compared CD + PRP vs. CD alone. Double-blinded; 4.5–6-year follow-up. Outcomes: HHS, modified Kerboul angle on MRI, stage progression, and Kaplan-Meier survivorship.	40 patients (53 hips) Mean age 36.6 years old	• Final HHS: 86.5 ± 11.6 (PRP) vs. 72.2 ± 10.1 (control); p < 0.001. • Modified Kerboul angle decreased in Group A (−12.4°) but increased in Group B (+10.6°); between-group difference p = 0.04. • Stage progression: 24% (PRP) vs. 43% (control); p = 0.025. • Survivorship free from THA: 92% vs. 78% (p = 0.01); free from femoral head collapse: 84% vs. 68% (p < 0.001). • PRP after CD provides significant pain relief, better functional outcomes, and enhanced hip survivorship compared with CD alone.
Martinot et al. (2020) [8]	Retrospective Case-Control Study	Compared CD alone vs. CD + non-concentrated bone marrow, or CD + BM + rhBMP-7. Primary outcome: native hip survival at 2 and 10 years. Secondary outcomes: HHS, PMA, UCLA, and Devane scores.	92 patients 83.6% male; mean age 41 yr	• Overall, 10-year hip survival 56.7% (95% CI: 47.0–68.4%); 2-year survival 64.4%. • Adding either BM or BMP improved 10-year hip survival (HR: 0.492; p = 0.035). • CD + BMP significantly improved 10-year hip survival vs. CD alone (HR: 0.356; p = 0.030); BM alone did not reach significance. • Kerboul angle > 60° and necrosis volume > 30% on MRI were independent risk factors for THA (both HR ≈12.5–13.0; p < 0.001).
Jindal et al. (2021) [46]	Systematic Review & Meta-Analysis	Analyzed 3 comparative studies of CD + BMAC in post-collapse ONFH. Primary outcomes were radiographic worsening and THA conversion.	270 total hips 196 hips treated with CD + BMAC	• XR progression in 39.8%; THA conversion in 39.3% (stage III) & 27.8% (stave IV) • No significant benefit of adding BMAC to CD over CD alone for progression or THA conversion • (BMAC does not appear to provide additional benefit over CD alone in post-collapse (stage 3–4) AVN; structural augmentation (tantalum rods, strut grafts) may be required.
Wang et al. (2024) [44]	Systematic Review & Bayesian Network Meta-Analysis	NMA of 17 studies comparing 6 regenerative strategies combined with CD: ABG, ABG + BMAC, BMAC, FVBG, MSCs, and PRP. Outcomes ranked by; primary metrics: ONFH progression rate and THA conversion rate.	17 studies; 1,019 hips	• Only MSCs and BMAC significantly outperformed CD alone in preventing ONFH progression: MSCs (OR: 0.098; SUCRA = 0.705) and BMAC (OR: 0.27; SUCRA = 0.431). • Only MSCs and BMAC significantly outperformed CD alone in preventing THA conversion: MSCs (OR: 0.062; SUCRA = 0.902) and BMAC (OR: 0.32; SUCRA = 0.511). • ABG, ABG + BMAC, FVBG, and PRP showed no significant advantage over CD alone. • MSCs ranked first among all six therapies for both outcomes; BMAC ranked second. • Evidence quality was generally low-to-moderate
Sun et al. (2026) [58]	Retrospective Cohort Study	72 patients with precollapse ONFH divided equally into: HBO alone (n=24), CD alone (n=24), or HBO + CD combination (n=24). Outcomes: OHS, SF-36 QoL, patient satisfaction, and Ficat radiographic progression at 12 months.	72 patients (24 per group) Mean follow-up ≈19.6 months Ficat stage II	• All three treatments produced significant improvements in OHS and SF-36 scores (all p < 0.05). • OHS was highest in the combination group (38.75 ± 2.88) vs. HBO (37.00 ± 3.20) and CD (36.20 ± 4.40); near-significant between-group difference (p = 0.05). • SF-36 scores differed significantly across groups (p = 0.0038), with the combination group highest. • XR no-progression rates not statistically significant: 90% (combination), 80% (HBO), 75% (CD) • HBO monotherapy showed efficacy comparable to CD alone; combined HBO + CD provided additive benefit in functional outcomes and quality of life.

Abbreviations:OHS Oxford Hip Score, PRP platelet-rich plasma, QoL quality of life, rhBMP-7 recombinant human bone morphogenetic protein-7, SF-36 Short Form-36

Regenerative Biologic Therapies

The most extensively studied biologics are BMAC and MSCs (Table 7). A growing body of evidence suggests that regenerative biologic augmentation may improve femoral head survivorship and reduce progression rates [48–51]. The data thus far is heterogeneous, making it difficult to make definitive conclusions about their efficacy. Larger randomized controlled trials are needed to clearly delineate the role of regenerative biologic therapies.

Bone Marrow Aspirate Concentrate (BMAC)

Bone Marrow Aspirate Concentrate (BMAC) is one of the most studied adjuvants with core decompression in the management of early-stage (ARCO I or II) ONFH. The addition of BMAC to CD is theorized to enhance osteogenic potential by providing functional mesenchymal stem cells as well as endothelial cells to restore biological conditions capable of healing [48–49]. Current literature is heterogenous regarding BMAC, with studies suggesting it may be useful in preventing further collapse in early stages of ONFH but losing its efficacy in later stage ONFH.

A meta-analysis by Tang et al. demonstrated that BMAC was more efficacious than BMP, PRP, or adjuvant osteoblasts in reducing both disease progression (RR = 0.47, p < 0.001) and THA conversion rates (RR = 0.41, p = 0.005) in younger patients when compared to CD alone [48]. Further, a retrospective review of 92 hips by Marinot et al. demonstrated improved 10-year survivability without conversion THA with addition of BMAC + BMP-7 to CD when compared to CD alone (HR = 0.356, p = 0.03). There was no difference in 10-year survivability with the addition of BMAC alone to CD (HR = 0.567, p = 0.13). However, hip survivability was similar between BMAC alone and BMAC + BMP-7 groups with CD at 10 years (HR = 0.612, p = 0.27) [52].

A retrospective review by Bootanapibul et al. evaluated the rate of collapse in hips with ONFH that were treated with either CD alone versus CD + BMAC based on pre-operative modified kerboul angles [53]. The authors demonstrated that CD + BMAC had significantly lower rates of collapse in early stages of ONFH with grade I (0%, 0/12 hips versus 33%, 3/9 hips) and grade II (12%, 2/16 hips versus 54%, 7/13 hips) modified kerboul angles compared to CD alone. However, there was no difference in rate of collapse in grade III and grade IV modified kerboul angles between CD alone and CD + BMAC.

A systematic review by Jindal et al. found that in the management of later-stage ONFH with femoral head collapse, CD combined with BMAC conferred no statistically significant advantage over CD alone in delaying radiographic progression (39.8% 66/166, OR = 1.41) or THA conversion (39.3% 70/178, OR = 0.92), suggesting that the regenerative capacity of BMAC is insufficient to overcome established subchondral collapse [49].

Mesenchymal Stem Cells

While various studies have investigated the role of stem cells in the management of pre-collapse ONFH, the data is heterogeneous and difficult to draw reliable conclusions. A meta-analysis by Zhang et al. demonstrated that, despite heterogeneous study designs, adjuvant MSCs with CD appear to lower VAS hip pain (mean difference= -10.88, p = 0.003) and increase HHS scores (mean difference = 5.59, p = 0.01) compared to CD alone. Authors also demonstrated lower rates of progression to femoral head collapse compared to CD alone (27.0% versus 51.4%, RR = 0.51, p = 0.0002) and THA conversion (22.5% versus 42.3%, RR = 0.49, p = 0.001) when bone marrow stem cells were added to CD when compared to CD alone [51]. Smaller studies demonstrate variable results but frequently report similar methodologies of harvesting bone marrow aspirate, isolating mesenchymal stem cells and osteoblasts ex vivo, followed by reinjection into areas of necrosis in combination with CD [54–56]. Overall, larger randomized controlled trials are needed to more reliably demonstrate the efficacy of mesenchymal stem cells in combination with core decompression.

Platelet-Rich Plasma (PRP)

Platelet-rich plasma (PRP) has emerged as a promising adjunctive therapy, particularly in early-stage disease. Its therapeutic effects are understood to operate through three principal mechanisms: inducing angiogenesis and osteogenesis, inhibiting inflammatory reactions in necrotic lesions, and preventing apoptosis [47]. Several early studies show promise in preventing femoral head collapse when PRP is used in conjunction with core decompression, but meta-analyses remain inconclusive [8, 47, 57]. In a prospective study of 53 hips, Aggarwal et al. demonstrated improved HHS, pain scores, and functional scores at 5-year follow-up with PRP + CD versus CD alone in early-stage ONFH. The authors also demonstrated significantly less progression of radiographic collapse of the femoral head in PRP + CD compared to CD alone (24% versus 46%, p = 0.025) at 5-year follow-up [58]. This study suggests that the addition of PRP both improves clinical and radiographic outcome measures, though the study population is small. Larger scale investigation is warranted to confirm the effectiveness of adding PRP to CD for pre-collapse ONFH management.

Bone Morphogenetic Proteins (BMPs)

Bone morphogenic proteins (BMPs) have demonstrated osteoinductive properties in preclinical studies; however, clinical evidence remains limited [52, 59]. BMPs assist in bone formation by stimulating osteoblasts and being pro-angiogenic. A recent study by Marinot et al. suggests there may be additional benefit to adding rhBMP7 to CD+BMAC for early-stage ONFH. Authors demonstrated improved 10-year survivability without conversion THA with addition of BMAC + BMP-7 to CD when compared to CD alone (HR = 0.356, p = 0.03) [52].

Hyperbaric Oxygen Therapy

Hyperbaric oxygen therapy (HBOT) has been proposed to enhance angiogenesis, reduce intraosseous pressure through enhanced venous outflow, and increase tissue oxygenation [60–61]. Several studies have suggested that HBOT has a positive effect on patient reported outcome measures and radiographic parameters [62–63]. Studies thus far have suggested HBOT is more effective in addition to CD, rather than as an isolated treatment. Heterogeneity in study design has prevented any definitive conclusions in meta-analyses on the effectiveness of HBOT [64].

Intraosseous Bisphosphonates

Bisphosphonates are a cheaper alternative to many adjuvant therapeutic options, and have been extensively studied in oral, subcutaneous, and intravenous routes. Animal studies have demonstrated good bioavailability of bisphosphonates with intraosseous administration, but few human studies have been completed [65]. Kumar et al. evaluated intraosseous ibandronate in a small case series of six hips, demonstrating satisfactory clinical outcomes and no radiographic progression at one year follow up [66].

Discussion

Osteonecrosis of the femoral head is a challenging condition that disproportionately affects young adults, placing them at risk of early progressive joint destruction and a high degree of morbidity at a young age [1]. While arthroplasty outcomes are excellent in late stage ONFH, early onset of disease and concerns for need of future revision arthroplasty continue to be of high concern [45–46]. Joint preservation strategies are best implemented in early-stage disease, prior to the collapse of the femoral head. While core decompression is the mainstay of hip preservation surgical intervention, there have been substantial efforts to identify adjunctive treatments that can improve outcomes and decrease progression of disease.

Among the treatment strategies reviewed, biologic augmentation combined with core decompression represents the area of greatest recent advancement and therapeutic promise. Regenerative biologic therapies appear to confer their benefit by restoring angiogenic and osteogenic potential within the necrotic zone. Biologic recovery depends fundamentally on the structural integrity of the subchondral plate remaining intact, potentially explaining why the benefit of these adjuncts diminishes once collapse has occurred [49]. To prevent the progression of femoral head collapse, early surgical intervention with biologic augmentation is more favorable than delayed intervention. Of the regenerative orthobiologic options available, BMAC is the most studied and supported adjuvant to core decompression, but it is most effective in pre-collapse ONFH (ARCO stage I and II) [47–49]. Other regenerative orthobiologics with promising early findings include BMPs and MSCs, though more high-quality studies are necessary to elucidate their true effect [8–51]. There has been no substantiative evidence to support the benefits of any adjuvant orthobiologic option in the treatment of late-stage ONFH when femoral head collapse is already present.

Core decompression continues to be the mainstay surgical intervention in the management of early-stage ONFH (ARCO I or IIa) [27–28]. However, a growing body of evidence suggests supplemental surgical interventions may improve hip survivability by decreasing femoral head collapse progression or delaying conversion THA. Hip arthroscopy and bone grafting as adjuvants to core decompression are the most well supported. Hip arthroscopy addresses concomitant intra-articular pathology and helps directly visualize the site of necrosis for decompression [32]. Bone grafting has several different options, including vascularized versus non-vascularized or autograft versus allograft. While no evidence suggests any particular bone graft option is superior to others, the concept of bone grafting suggests improved results when added to core decompression [36–37].

While pharmacologic agents (like bisphosphonates and denosumab) and interventions (like HBOT) offer meaningful support for collapse prevention, their role is only adjunctive and should not replace or delay surgical management in operative candidates. As implant technology and bearing surfaces continue to improve outcomes for THA in younger patients, the risk-benefit calculus for complex joint preservation procedures merits ongoing reassessment on a case-by-case basis.

Despite the growing body of research in ONFH, the clinical application of many emerging therapies remains constrained by the quality and consistency of the underlying evidence. The majority of studies evaluating biologic augmentation, bone grafting, and adjunctive procedures are single-center, observational, and have notable heterogeneity in patient selection and outcome reporting which limit definitive comparative conclusions. Future research should prioritize multicenter randomized controlled trials with standardized outcome reporting, ideally stratified by ARCO stage and lesion characteristics. Advances in predictive modeling may further refine patient selection and help direct biologic augmentation toward those most likely to benefit. Progress in these areas will be essential to closing the gap between the current evidence base and the standardized, patient-specific treatment algorithms this condition requires.

Conclusion

Osteonecrosis of the femoral head remains a challenging condition with a variable natural history. A variety of pharmacological and surgical interventions are available in the treatment of early stage ONFH with the goal of preventing femoral head collapse. Core decompression continues to serve as the foundation of treatment in pre-collapse disease, with emerging evidence supporting biologic augmentation strategies, with BMAC having the most support in the literature. Supplemental surgical operations such as hip arthroscopy and bone grafting have also demonstrated promising results when combined with core decompression. The current body of literature is limited by the heterogeneity in study design, lack of high-level randomized controlled trials, and variability in outcome measures and follow-up duration. While recent advances are promising, further high-quality research should focus on standardization of therapies and comparative effectiveness trials to determine long term outcomes, establish standardized treatment algorithms, and optimize patient care.

Key References

• Mont M, Salem H, Piuzzi N, Goodman S, Jones L. Nontraumatic Osteonecrosis of the Femoral Head: Where Do We Stand Today?: A 5-Year Update. JBJS Am. 2020;102(12):1084-1099. 10.2106/JBJS.19.01271

  • ○  Key summary article on general ONFH topic defining epidemiology, pathophysiology, diagnostics, and treatment option summaries
• Chow S, Rainey J, Mont M, Jones L, Goodman S. The Role of Cell Therapies for the Treatment of Osteonecrosis of the Femoral Head. J Arthroplasty. 2025;40(10S1):S46-S50.10.1016/j.arth.2025.06.018

  • ○  Key summary article about the most important regenerative orthobiologic therapies as adjunctives to core decompression, BMAC and MSCs
• Mont MA, Carbone JJ, Fairbank AC. Core decompression versus nonoperative management for osteonecrosis of the hip. Clin Orthop Relat Res. 1996;324:169-178. 10.1097/00003086-199603000-00020.

  • ○  Landmark article highlighting the benefit of core decompression as management strategy in early stages of ONFH compared to non-operative treatment
• Domb B, Kahana-Rojkind A, Shah P, Rana K, Becker N, Quesada-Jiminez R. Concomitant Hip Arthroscopy Is Associated With Improved Outcomes of Core Decompression for Osteonecrosis of the Femoral Head: A Systematic Review. Arthrosc Sports Med Rehabil. 2025;7(5). 10.1016/j.asmr.2025.101225

  • ○  Key summary article identifying the role of hip arthroscopy as a beneficial adjunctive to core decompression, resulting in improved PROMs and femoral head survivorship
• Hu L, Deng X, Wei B, Wang J, Hou D. Comparative analysis of surgical interventions for osteonecrosis of the femoral head: a network meta-analysis of randomized controlled trials. J Orthop Surg Res. 2023;18(1). 10.1186/s13018-023-04463-4

  • ○  Key summary article evaluating the variety of different bone grafting options. It demonstrated no clear superiority amongst the bone graft options, but demonstrated a benefit of all bone grafting options when used as adjuncts compared to core decompression alone

Author Contributions

Authors TS and ND performed literature review and manuscript preparation. Author JG organized search strategy for authors, discussed structure of manuscript, and reviewed manuscript with appropriate edits.

Funding

No funding was received.

Data Availability

No datasets were generated or analysed during the current study.

Declarations

Human and Animal Rights and Informed Consent

This article does not contain any studies with human or animal subjects performed by any of the authors.

Competing interests

Author JG is a committee member for AOSSM and on the editorial board for Journal of Arthroscopy and OJSM.