Clinical result of structural augmentation with cannulated bone screws for the treatment of osteonecrosis of the femoral head

Abstract

Objective:  To evaluate the results of treatment of osteonecrosis of the femoral head by structural augmentation through a routine core decompression procedure combined with insertion of cannulated bone screws incorporating autogenous bone graft and biomaterial containing decalcified bone matrix.

Methods:  From February 2002 to February 2005, 31 patients (33 hips) with femoral head necrosis were treated in our hospital using insertion of cannulated bone screws incorporating autogenous bone graft. There were 18 men and 13 women with an average age of 37 years (range, 27–49). The Steinberg classification was stage I for 20 hips (61%) and stage II for 13 hips (39%). Clinical and radiographic evaluations were performed on all patients. The patient's satisfaction was also assessed.

Results:  All 31 patients (33 hips) were retrospectively studied after a mean follow‐up of 38 months (range, 18–48). The average Harris hip score was 76 before surgery and 91 at the final follow‐up. All patients stated that they were satisfied and had significantly reduced pain. According to the Harris hip score system, 21 cases were excellent, 8 good and 2 fair. No complications, such as wound infection, subtrochanteric fracture, neuropathy and deep vein thrombosis, were found.

Conclusion:  Structural augmentation using the insertion of cannulated bone screws incorporating autogenous bone graft is an effective option for Steinberg I–II stages of femoral head necrosis. Further study is needed to confirm mid‐ and long‐term results.

Introduction

Osteonecrosis of the femoral head (ONFH) occurs predominantly in younger patients, and without treatment most of them suffer from collapse of the femoral head and painful arthritis of the hip joint within several years. The ideal goal of early treatment is to delay or arrest the progression of the disease before articular collapse occurs. Over the past ten years various types of femoral head‐preserving surgery, such as core drilling, bone grafting, and osteotomy have been reported ¹ , however there is uncertainty regarding the optimal surgical procedure for the treatment of this disease. In order to prevent mechanical failure of the subchondral bone and articular collapse, cannulated bone screws incorporating autogenous bone graft and biomaterial containing decalcified bone matrix (DBM) were used to treat early ONFH in this study. The aim was to help to get a balance between bone resorption and formation, strengthen the structural mechanics of the femoral head, and provide structural support to the articular cartilage.

Materials and methods

Case data

From February 2002 to February 2005, 31 patients (18 men and 13 women with 33 hips), were treated with a routine core decompression procedure combined with the insertion of cannulated bone screws incorporating autogenous bone graft and biomaterial containing DBM. Eleven cases (13 hips) were corticosteroid‐induced, seven (7 hips) due to excessive alcohol consumption and thirteen (13 hips) of unknown cause. Systemic lupus erythematosus was the reason for the use of steroids in four patients (six hips). The evaluation of preoperative radiographs indicated that 20 hips (61%) were stage I and 13 hips (39%) stage II according to the Steinberg classification. All patients complained of pain in the inguinal region or hip joint. Their preoperative average score was 76 points (range, 39–81) according to the Harris hip score system ² . The severity of ONFH and the necrotic area of the femoral head were evaluated with radiographic views, magnetic resonance imaging (MRI) or computed tomography (CT).

Operative technique

The patient was positioned on a fracture table to permit image intensification during the procedure, and the contralateral hip placed in a flexion and abduction position. The affected limb was positioned in extension and sufficient internal rotation to neutralize anteversion of the femoral neck. Then, the involved hip was prepped and draped in a routine fashion. Under image intensification, a 9‐mm entry hole was drilled and a guide wire inserted deep into the subchondral bone in the center of the lesion. A 9 mm cannulated drill was advanced to two‐thirds of the intended depth (within approximately 30 mm of the articular surface) following the guide wire. The remaining depth of the canal, to the subchondral plate within approximately 5 mm of the articular cartilage, was carefully completed using an 8 mm drill and advancing the tap slowly to create the intended thread and the area for insertion of the interference screw and osteoinductive material into the depth of necrotic lesion. Under image intensification, a bone interference screw incorporating autogenous bone graft and biomaterial containing DBM was driven into the full depth of the canal thus created. The remainders of both tracts from the lateral femoral cortex to the back end of the screw were packed with a combination of corticocancellous bone chips. The patients were instructed not to weight‐bear for six weeks. After six weeks, partial weight‐bearing with crutches was allowed for three to ten weeks.

Case example

Case one

A 49‐year old man presented with left groin pain and a painful range of movement for two years. CT scan and radiographic study showed evidence of osteonecrosis of the left hip in the pre‐collapse (Steinberg II) stage with cystic degeneration in the femoral head region. The procedure of cannulated bone screw insertion incorporating autogenous bone graft and DBM was performed on this patient (Fig. 1). The patient was satisfied with the result one year after surgery; the cannulated bone screw was slowly substituted by some autologous bone without collapse of the head.

Figure 1.

A 49‐year old male patient presented with left groin pain and painful range of motion for two years. (a) Radiograph and (b) CT scan show evidence of osteonecrosis of the left hip in the pre‐collapse stage (Steinberg II) with cystic degeneration in his femoral head region. (c) The procedure of cannulated bone screw insertion incorporating autogenous bone graft and DBM was performed. (d) One year later, the cannulated bone screw was slowly substituted by some autologous bone and no collapse of the head had occurred.

Case two

A 36‐year old woman complained of right hip pain for six months with intensification in the last ten days. No abnormality was detected on radiographic examination, however osteonecrosis of right femoral head was identified by MRI (Steinberg stage I). At 2.5 years after index surgery, the patient remained clinically asymptomatic, and exhibited a painless and full range of movement (ROM) of the hip. A follow‐up radiographic study showed the screws in situ, with evidence of a bone healing response to the osteoinductive material and no collapse (Fig. 2).

Figure 2.

A 36‐year old female patient presented bilateral hip pain for eight months, the pain had intensified in the previous two weeks. (a) Radiograph shows evidence of osteonecrosis of both hips in the pre‐collapse stage. (b) Half a year later, a follow‐up radiograph shows screws in place in what appears to be a stabilized lesion with evidence of a bone healing response to the osteoinductive material. (c) One and a half years later, the cannulated bone screw had been partly substituted by autologous bone. (d) Three years after the operation, the cannulated bone screw had been substituted entirely in the right hip and partially on the left side, and no collapse of the bilateral head had occurred.

Results

All 31 patients (33 hips) were retrospectively studied for a mean follow‐up of 38 months (range, 18–48). Pain disappeared in 21 patients, who needed no analgesics, and was significantly alleviated in eight patients who required analgesics occasionally. Pain was relieved to some extent in the remaining two patients after surgery, but regular analgesics were necessary.

Furthermore, in all cases, the ROM was improved to normal or almost normal, especially in whose pain was greatly relieved. According to the Harris hip score system, the average score was 91 points post operation, and the scores were graded as excellent in 21, good in 8 and fair in 2 cases. No complications such as wound infection, subtrochanteric fracture, neuropathy and deep vein thrombosis (DVT) were found.

Discussion

Over the past few decades osteonecrosis has represented a source of frustration for orthopedic surgeons attempting to reach the ideal surgical outcome: arrest of progression of the disease and its devastating consequences. Approximately 10% to 12% of all hip arthroplasties performed in the USA are a consequence of avascular necrosis and the ensuing articular collapse and osteoarthrosis ³ . The problem is further compounded by the fact that the patients requiring total hip replacement due to avascular necrosis remain young.

Relationship between histopathology changes in ONFH and opportunity for surgery

The histopathology of osteonecrosis of the hip is fairly constant. The areas of necrotic bone and infarction trigger an inflammatory response with reactive hyperemia, neovascularization, and deposition of vascular fibrous tissue in the periphery of the lesion. The repair process is characterized by a weak balance between bone resorption and bone deposition (creeping substitution). New woven bone is laminated onto necrotic trabeculae with resorption of the necrotic bone. The combination of irregular areas of bone deposition and bone resorption can be appreciated in plain films as ‘fragmentation’ or areas of sclerosis interspersed with areas of increased density. The bone resorption rate is greater than the bone formation rate and may also play a role in the development of microfractures in the depth of the femoral head. The imbalance between removal of necrotic bone and calcification of vascularized immature bone creates a ‘vulnerable gap’ where the new woven bone fails beneath the subchondral plate. The final pathway, however, remains unchanged: gross subchondral mechanical failure will ultimately lead to articular collapse and subsequent osteoarthritis.

After structural failure of the subchondral bone, most patients will eventually require a total hip replacement. The fate of the primary arthroplasty in these young, active patients is cause for concern. Because their life expectancy exceeds the ‘life span’ of the biomaterials used in joint replacements, they are likely to require at least one revision surgery in the future. The ideal goal of any ‘early’ treatment is to delay or arrest the progression of the disease before articular collapse. Consequently, the most favorable time to intervene is early in the history of the disease, before mechanical failure of the subchondral bone and articular collapse (Steinberg stages I and II).

Similar techniques and literature review

The clinical outcome of core decompression and cortical bone grafting techniques depends mainly on the stage and extent of the preoperative lesion. Satisfactory clinical results have been reported only in pre‐collapse lesions ⁴ , ⁵ . In a retrospective literature review encompassing 42 reports with 2025 patients, Mont et al. found satisfactory clinical results in only 22.7% of the patients treated conservatively but in 63.5% of those treated with core decompression ⁶ . The rate of progression to a total hip replacement ranges from 31% to 57%. Vascularized bone grafts have been shown to incorporate more rapidly and predictably than non‐vascularized grafts ⁷ . In addition to structural support, vascularized bone grafts introduce a source of mesenchymal stem cells and a well‐defined vascular supply, which makes them a valuable alternative for patients with advanced stages of the disease. The technique, however, requires considerable technical expertise, the participation of two operating teams, and a prolonged surgical and rehabilitation time. It also involves increased morbidity associated with the donor site and the femoral neck itself.

In the past several authors have attempted to augment a standard core decompression procedure. Soucacos et al. tried to update core decompression with homogenous variant bone grafting for structural support of the femoral head ⁸ . Then Berend et al. ⁹ improved this technique. They utilized the core decompression channel for autogenous or homogenous fibular or ilial grafting. Though the short‐term result was excellent, in the long‐term the survival rate of the femoral head was low. Recently, Plakseychuk et al. retrospectively compared the results for 50 hips treated with a free vascularized fibular graft performed at the University of Pittsburg Medical Center, and 50 hips treated with a non‐vascularized fibular graft performed at the Kyungpook University Hospital in Korea ¹⁰ . After an average follow up of 5 years the mean Harris hip score had improved in 70% and 36% of hips respectively. However the rate of complications was higher in the vascularized group, including a subtrochanteric fracture, clawing of the great toe in three patients, and transient peroneal neuropathy in four. The rate of conversion to total hip arthroplasty was not mentioned.

Ciombor and Aaron compared the clinical outcomes in two groups of patients undergoing core decompression with the addition of DBM versus core decompression alone ¹¹ . Overall, there were 43 Steinberg stage II and 47 Steinberg stage III hips in the core decompression alone group and 27 Steinberg stage II and 20 Steinberg stage III hips in the core decompression plus demineralized bone graft group. Even though the initial results were promising, at long‐term follow‐up only an 11% improvement in hip survival was found in Steinberg stage II hips treated with core decompression and DBM compared with core decompression alone. There was no additional benefit observed in Steinberg stage III hips. The procedure included only the introduction and tamping of 6 to 7 ml of DBM into the core decompression tract. No structural support was added to the construct. Leali et al. reported a modified core decompression procedure combined with the insertion of two interference screws into the subchondral plate to provide structural support, and the use of osteoinductive material (i.e. demineralized bone matrix) in an effort to accelerate the bone healing process, but the document is only a case report without follow up of the patient ¹² .

Character of structural augmentation using cannulated bone screw incorporating autogenous bone graft

This article attempts to integrate several treatment principles for avascular necrosis. The elevated intraosseous pressure is relieved with a standard core decompression which also improves vascularity and relieves pain. The structural deficit is addressed with the use of bone screws. The bone healing process is accelerated with the incorporation of osteoinductive material into the drilled tracks in the depth of the subchondral bone ¹³ , ¹⁴ . The goal is to tip the balance of the creeping substitution process by accelerating bone healing while providing enough structural support to the articular cartilage. Although the purpose of the present article was to introduce the surgical technique and its rationale, the authors believe that long‐term stabilization of the lesion can be achieved with a reasonably simple procedure involving a short postoperative period with low morbidity. A prospective trial to assess mid‐ and long‐term clinical and imaging outcomes is currently underway.