Issue: November 2016
November 07, 2016
13 min read

Panel debates pathophysiology, treatment for osteonecrosis of the femoral head

Issue: November 2016
You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

Osteonecrosis of the femoral head remains a challenging condition in young patients and often results in total hip arthroplasty. I believe it is time for an Orthopedics Today Round Table on recent advances in our understanding of the pathophysiology, diagnosis, staging and treatment for this condition.

There has been recent chatter about core decompression, with backfill procedures and orthobiologics as preferred adjunctive or isolated treatment options compared with vascularized fibular grafts or osteotomies. Does hip arthroscopy play a helpful role in the treatment of osteonecrosis of the femoral head (ONFH)? What considerations are important to total hip arthroplasty (THA) longevity in this setting?

It was my distinct pleasure to invite this faculty panel for a candid and controversial discussion about this topic.

Dean K. Matsuda, MD

Roundtable Participants

  • Dean K. Matsuda, MD
  • Moderator

  • Dean K. Matsuda, MD
  • Marina del Rey, Calif.
  • Jay R. Lieberman, MD
  • Jay R. Lieberman, MD
  • Los Angeles
  • Michael A. Mont, MD
  • Michael A. Mont, MD
  • Cleveland
  • Shane J. Nho, MD, MS
  • Shane J. Nho, MD, MS
  • Chicago
  • Rafael J. Sierra, MD
  • Rafael J. Sierra, MD
  • Rochester, Minn.

Dean K. Matsuda, MD: What are the most important recent advances in our understanding of the pathophysiology of ONFH?

Jay R. Lieberman, MD: Multiple hypotheses remain to explain the development of ON of the hip, including ischemia, direct cellular toxicity and altered differentiation of mesenchymal stem cells, and there is no one hypothesis to date that explains the cell death and progression to collapse of the femoral head associated with ON. However, Matthieu Ollivier, MD, and his colleagues recently published a paper in the Journal of Bone and Joint Surgery related to patients with idiopathic ON of the hip. Seventy percent of patients with idiopathic ON of the hip had three of four anatomic anomalies. For example, a decreased femoral neck shaft angle less than 129°, increased femoral neck version greater than 17°, lateral center-edge angle of less than 32° and acetabular version less than 19°. Only 11% of control patients had these abnormalities. This study suggests anatomy may have an influence on the development of ON in patients with idiopathic ON of the hip and this clearly requires further study.

Michael A. Mont, MD: Multiple reports have recently been published on the association of genetic polymorphisms of enzymes and ONFH. Some of the genes that have been implicated in this include matrix metalloproteinase (MMP)-8, MMP-9, P-glycoprotein gene ABCB1, C3435T and G2677T/A in steroid-induced ONFH. For alcohol-associated ONFH, apolipoprotein (ApoA) B and ApoA1 have been implicated. Recent studies have also demonstrated an association with pro-inflammatory cytokines (IL-23, IL-1alpha, IL-10, TNF-alpha, etc.) and lipid metabolism alterations that could be relevant to ONFH pathophysiology. More work needs to be done in this field for a clearer understanding.

Matsuda: What are the top advances in the non-surgical treatment of ONFH?

Lieberman: There is limited data to demonstrate bisphosphonates, low-molecular weight heparins or protected weight-bearing have any real impact on preventing collapse of the femoral head. There are some retrospective studies showing that low-molecular weight heparins may help prevent collapse in patients with ON and hypercoagulability, but these studies include small numbers of patients who were not randomized. In a recent randomized trial, zoledronate was not effective in preventing collapse of the femoral head. I am not aware of any studies that show that protected weight-bearing alone is useful.

Rafael J. Sierra, MD: Although there is growing interest in the non-surgical treatment of ON, there have been few top advances (other than finding out what does not work) in this area during the past decade. Several studies have evaluated the role of observation and protected weight-bearing of the affected limb. One can conclude from the studies that larger lesions (greater than 25%), located in the weight-bearing zone of the femoral head tend to progress at an average of 3 years. Non-weight-bearing does not seem to affect the progression, so I usually allow my patients to weight bear as tolerated, unless they are doing it for pain relief.


Observation may be an option in asymptomatic lesions, but once the hip becomes symptomatic, some other form of management should be instituted. Pharmacologic treatments — the use of bisphosponates, prostaglandin analogs, anticoagulants and other non-pharmacological treatments, such as the use of extracorporeal shockwave therapy or hyperbaric oxygen — have been studied. The variability in disease characteristics and natural history of the disorder make it hard to positively conclude the effectiveness of these treatment options.

pre-collapse osteonecrosis of the femoral head
Shown is postoperative modified Dunn lateral view of right hip of 25-year old man with pre-collapse osteonecrosis of the femoral head. Initial arthroscopy permitted labral repair and assessment of articular cartilage followed by percutaneous core decompression, local bone grafting (harvested from trochanteric region), and backfill with calcium sulphate/phosphate self-hardening regenerative bone graft substitute (PRO-DENSE, Wright Medical Group). Arrow shows region of concurrently performed arthroscopic femoroplasty.

Image: Matsuda DK

There have been several level 1 studies evaluating these treatments with mixed results. For example, the role of bisphosphonates is controversial. There was marked early interest in their potential role in preventing collapse. However, a multicenter, randomized controlled trial found no difference between alendronate and placebo in preventing femoral head collapse and delaying THA. Charles Glueck, MD, and his colleagues reported on a select group of patients with ON and underlying coagulopathies with the use of a course of anticoagulants, such as low-molecular weight heparins, in order to delay progression of femoral head ON. Further research is needed to further clarify the benefits of this therapy. Interestingly, a randomized trial showed superiority of hyperbaric oxygen compared with compressed air in patients with symptomatic ON after 20 sessions or 30 sessions.

Outcomes led to improved pain and reduction in the need for THA. Similar results have been seen with extracorporeal shockwave therapy. Widespread use of these therapies is limited due to their availability and costs.

Matsuda: Which hip preservation surgeries do you prefer? Why and what are their indications?

Mont: For early stage ONFH, I usually perform multiple percutaneous drillings using a 3.2-mm to 3.4-mm drill, attempting to delay or prevent the need for THA. The success of this procedure is higher for smaller lesions. For example, when less than 15% of femoral head is involved or the Kerboul angle is less than 200. Some surgeons use core decompression with adjunctive treatments, such as bone grafting and/or the addition of bone morphogenetic proteins or mesenchymal cells. For larger lesions, bone grafting procedures can be done with possible supplementation with bone morphogenetic proteins, bone marrow, mesenchymal cells and/or corticocancellous bone grafts.

In the lightbulb procedure, a conduit is created from a cortical window in the femoral neck toward the necrotic area. The necrotic tissue is removed until only healthy bone remains and bleeding is visualized in all areas. This is then packed with various grafting materials as mentioned.

Shane J. Nho, MD, MS: My indication for hip preservation surgery is symptomatic avascular necrosis (AVN) without collapse. My preference is to perform an arthroscopic-assisted core decompression with a bone graft substitute. I prefer to use the PRO-DENSE bone graft substitute (Wright Medical Group), which is a calcium phosphate that provides good biomechanical strength, as well as biologic resorption and remodeling over several months. The AVN is generally difficult to localize on fluoroscopy. Therefore, using the arthroscope allows you to see the chondral lesion directly and to see the lesion directly. In addition, you can also use the arthroscope to look into the bone tunnel to see if the character of the bone changes in the region of necrosis.

Matsuda: What technical considerations do you deem important in the performance of core decompression?

Sierra: Typically, I have performed a simple hip decompression with the use of a 6-mm trochar that has an inner diameter of around 3.5 mm. Using no power, the trochar is hammered from the lateral cortex of the femur distal to the vastus ridge into the femoral head under fluoroscopic guidance. This type of decompression has shown to be effective in helping with pain induced from femoral head ON. Not only is it an effective way of doing a decompression, I find it to be easy and find it gives me the opportunity to inject fluid or bone graft into the site.

The morbidity of the procedure is low. Because of the smaller diameter, patients are allowed to immediately weight-bear as tolerated; this is beneficial in bilateral hip cases. Care should be taken, as with other decompressions, to avoid starting below the level of the lesser trochanter to prevent fracture. Once the trochar is in place in the femoral head, the surgeon can inject their preferred mix into the head. I inject a mix of 6 mL of platelet-rich plasma (PRP) and if possible, 12 mL of bone marrow concentrate into each femoral head. Then I inject a few milliliters of DBX demineralized bone matrix putty (DePuy Synthes) to plug the tract, but I do not backfill the entire tract. That is usually done with some cancellous bone from the tract, as the trochar can be slanted slightly in the tract and impacted back and forth several times to fill the canal with cancellous bone from the walls of the initial tract.


Nho: I use a single large bore tunnel centered over the AVN lesion. I will also use a currette to ensure complete removal of necrotic bone. I will backfill the core with PRO-DENSE bone graft substitute to minimize risk of bone collapse or fracture.

Matsuda: In the orthopedic literature, some researchers have reported using pre-collapse as the threshold for some hip preservation procedures whereas others use collapse less than 2 mm. What do you use and why?

Mont: Some of the post-collapse lesions (less than 2 mm) have intact cartilage amenable to the preserving procedures, which has been performed for pre-collapse lesions. Unfortunately, this has universally poor results. Many patients who have pre-collapse lesions have subtle areas of collapse and are not amenable to core decompression or percutaneous drilling. In patients with less than 2-mm collapse, hip preservation procedures typically are not used because the results have not been optimal for grafting or vascularized fibular grafting procedures.

Lieberman: My femoral head preservation procedure of choice at this time is a core decompression (single-large core tract and approximately 9 mm to 11 mm in size) with debridement of the necrotic bone in the femoral head and bone grafting with cancellous bone graft from the greater trochanter and concentrated stem cells obtained via bone marrow aspirate. A demineralized bone matrix putty is used to seal the core tract. In general, I do not perform this procedure if there is a crescent sign, any evidence of collapse on plain radiographs or the patient has significant limitation in internal rotation of the hip because in my experience, head preservation procedures in this group do not lead to satisfactory results. The lack of randomized trials has not allowed us to determine the optimal treatment for ON.

Matsuda: What role, if any, does hip arthroscopy play in the staging and treatment of ONFH?

Nho: I have not used arthroscopy for staging. I rely on the plain radiographs and MRI to ensure that there is no collapse. In some cases, I have identified a small AVN lesion prior to hip arthroscopy for a labral tear. My treatment algorithm has been to perform an intra-articular injection to determine how much of the pain is from the chondrolabral injury compared with AVN. If the patient experiences significant symptomatic improvement after the injection, then I will proceed with hip arthroscopy for labral and femoroacetabular impingement (FAI) pathology alone. If there is only partial relief of pain, then my thought is that some of the symptoms may be coming from the AVN, and I will treat simultaneously. With that said, there are few patients who have both pathologies.

Sierra: I personally do not see a role for hip arthroscopy in the diagnosis of ON. A MRI or CT scan can adequately stage the disease process and looking at the cartilage adds very little to the decision whether I will go ahead with preservation or not. However, there are patients who have radiographic features of FAI with labral and cartilage pathology that may present with ON. This type of patient is a diagnostic dilemma. In the past, I have treated such patients with simultaneous hip decompression and hip arthroscopy to manage the labral and cartilage pathology, but today, I favor treating the ON first with a decompression. If the patient remains symptomatic and has no evidence of progression of the ON on radiographs, then I proceed with hip arthroscopy for treating the labral and cartilage damage. About half of the patients will not need hip arthroscopy. I do tell them that there are case reports of patients undergoing hip arthroscopy that have developed ONFH, and that is a risk.

Matsuda: What role do you think orthobiologics, such as PRP, bone morphogenetic proteins or stem cells, play in the treatment of ONFH?

Sierra: There are now several randomized controlled trials that demonstrate improved pain and improved survivorship from THA when a hip decompression is augmented with bone marrow concentrate compared to a hip decompression alone. Some of these papers have also shown a decrease in the size of the femoral head lesion, but I personally have seen just a few examples of that. The injected cells have potent anti-inflammatory properties and may help in modulating the immune response within the femoral head that leads to improvements in pain and function, but I doubt that the injection will eventually heal the lesion in the majority of hips.


Using THA as an endpoint, our most recent, unpublished, prospective trial demonstrated an 89% success rate at a minimum of 2 years when injecting bone marrow derived mesenchymal stem cells into the femoral head in patients with early Steinberg stage 1 or stage 2 hips. Our results were highly dependent on the number of nucleated cells that were obtained from the iliac crest prior to injection. We believe that in certain patients, an alternative source of cells may be needed as the bone marrow may be an inadequate source in some patients with ON.

Lieberman: Unfortunately, there are no appropriately powered randomized trials evaluating biologic agents as bone graft substitutes for grafting of the femoral head. PRP has not been shown to be osteogenic in laboratory studies, so I would not recommend it as the sole bone graft substitute combined with core decompression. Intuitively one would think bone morphogenetic proteins could enhance bone repair in the femoral head, but there is limited data on the efficacy of bone morphogenetic proteins for treatment of ON. In addition, recombinant bone morphogenetic proteins are expensive, and it would be an off-label use of this protein. Concentrated stem cells harvested via bone marrow aspirate can be used to bone graft the femoral head. These data look promising, but stem cells combined with core decompression have not been assessed in randomized controlled trials.

Matsuda: If orthobiologics prove to be beneficial, do you predict that the role of non-vascular grafts or tantalum porous rods may play a bigger role in the future?

Mont: Non-vascularized bone grafts with biologics may have a role in the treatment of ONFH. We have reviewed 33 patients (39 hips) who received non-vascularized bone grafting for ONFH with supplementation osteogenic protein-1. No reoperations were performed in 25 of 30 small and medium lesions. Patients who had Ficat stage 2 lesions did not receive further intervention in 18 of 22 cases. More studies are needed in this direction to define the role of non-vascularized bone grafts, especially with the use of adjunctive growth factors (ie, bone morphogenetic proteins) or cell-based therapies. Based on the literature, tantalum rods do not have a proven role in the treatment of ONFH.

A retrospective study of 23 ONFH that underwent core decompressions with adjunction of tantalum rods reported a survival rate of 44% at a mean follow-up of 1.45 years. Few studies report on the use of this adjunctive treatment of ONFH. Overall, there is not enough supportive long-term follow-up data to support the use of this procedure.

Nho: I would predict that a structural scaffold will be beneficial if there are improvements in orthobiologics to enhance osteoinduction and osteoconduction. With access to autologous bone graft, it seems feasible with improvement in osteoinduction and osteoconduction that a non-vasularized graft should allow better treatment of AVN.


Matsuda: What are your indications for choosing THA over hip preservative surgery? What are your implant considerations?

Lieberman: In general, I would not recommend a hip preservation procedure in patients with collapse of the femoral head. If a patient is 50 years of age or older with extensive involvement of the femoral head, it may be more judicious to consider THA because the results are reliable. When performing a total hip replacement, I prefer a proximally coated, cement-less femoral stem and a cement-less acetabular component with a highly crosslinked polyethylene liner. Since most patients with ON are young, I generally use a ceramic femoral head.

One pearl regarding the preparation of the acetabulum is that the sclerotic bone typically associated with osteoarthritis of the hip is not present. Therefore, one should be gentle when performing the acetabular reaming and should not hesitate to use screws in the cup to enhance fixation if there are any concerns regarding the quality of the bone.

Sierra: Independent of age, THA is the most predictable option in patients with advanced ONFH. Once subchondral collapse occurs (Steinberg stage 3), joint preservation becomes unpredictable in my hands. In the young patient, I have tried alternative procedures in an effort to avoid THA. The decision to proceed with one or the other depends on the size of the lesion, amount of collapse, symptoms and etiology of the ON. Patients with less symptomatic, post-traumatic lesions that are fairly small (smaller than 25%) and have minimal collapse (less than 2 mm) with no evidence of acetabular cartilage damage may be the best candidates for attempting a head preserving procedure.

A trap-door procedure or elevation of the segmental bone collapse from within the decompression tract and followed by bone grafting has been successful in patients with similar characteristics. If THA is the best choice, then an uncemented THA using a 36-mm ceramic head on highly crosslinked polyethylene is the most common construct I use. I tend to use high-offset stems in all patients with an underlying diagnosis of ON, as our joint arthroplasty registry demonstrates that ON is an independent risk factor for dislocation. I would only consider a cemented stem in the older patient who is older than 70 years of age with ON, but this is a rare instance as most of these patients are young.

Disclosures: Lieberman reports he is a paid consultant for DePuy Synthes and receives stock or stock options from Hip Innovation Technology. Matsuda reports he receives royalties for intellectual property from Zimmer Biomet and Smith & Nephew. Mont reports he is a paid consultant for DJ Orthopaedics, Johnson & Johnson, Merz, Orthosensor, Pacira, Sage Products, Stryker, TissueGene and U.S. Medical Innovations; receives research support from DJ Orthopaedics, Johnson & Johnson, National Institutes of Health (NIAMS & NICHD), Ongoing Care Solutions, Orthosensor, Stryker and TissueGene; and receives IP royalties from Microport and Stryker. Nho reports he receives research support from Allosource, Arthrex, Athletico, DJ Orthopaedics, Linvatec, Miomed, Smith & Nephew and Stryker; receives publishing royalties, financial or material support from Springer; receives IP royalties from Ossur; is a paid consultant for Stryker and Ossur; and is a board or committee member of the American Orthopaedic Society for Sports Medicine and the Arthroscopy Association of North America. Sierra reports he receives royalty payments for the Perfuse system and G7 Acetabular System from Zimmer Biomet.