Keywords

Introduction

Untreated non-traumatic avascular necrosis (AVN) of the femoral head results in a very poor outcome, due to the high percentage of collapse of the head, ranging from 70 % to 100 % of cases, depending on the stage of the disease [9, 26, 29, 38, 42].

Ficat and Arlet described a classification system based on standard, radiographs, consisting of four stages (I to IV) to be applied to symptomatic patients [10, 11]. To this staging system, Hungerford added a stage “0”, pre-clinical and pre-radiographical. The classification of Ficat was later modified by several authors, based on magnetic resonance imaging. In 1992, the Association Internationale de Recherche sur la Circulation Osseuse (ARCO) proposed a classification based on the staging system of Ficat and Arlet and including both the quantification and the location of the involvement (Table 1). Mont and Hungerford, in an overview of non-traumatic avascular necrosis of the femoral head, suggested that this staging system be adopted as the international standard classification of this disease.

Table 1 The Ficat-Arlet staging system modified according to the Association Internationale de Recherche sur la Circulation Osseuse
Full size table

The radiographic grading of the Japanese Investigation Committee illustrates the location and extent of the avascular changes and a classification similar to that of Ficat and Arlet (Fig. 1).

Fig. 1
figure 1

Japanese Institution Committee radiographic classification of AVN of the Femoral Head (From: H. Matsusaki et al. in the Arch Orthop Traum Surg (2005)125;95–101)

Full size image

In an effort to improve the prognosis of avascular necrosis, several surgical procedures have been proposed, aimed at preventing the collapse of the subchondral bone and at slowing the progress of degenerative changes.

These procedures can be grouped, depending on the desired goal, as follows:

  1. 1.

    Procedures aimed at reducing intramedullary pressure;

  2. 2.

    Procedures aimed at moving the necrotic area out of the weight-bearing zone;

  3. 3.

    Procedures aimed at restoring the bone supply in the necrotic area.

The efficacy of these procedures has been variable, with reported success rates ranging between 60 % and 80 % at the time of short-term and mid-term follow up.

Procedures Aimed at Reducing Intramedullary Pressure

Arlet and Ficat first recognised the therapeutic effect of core decompression on AVN of the femoral head and described its rationale. Observing the raised intramedullary pressures and retarded venous sinusoidal drainage in the proximal femur, they developed the hypothesis that vascular stasis and prolonged ischaemia were factors involved in the pathogenesis of AVN [1, 10, 11]. According to Ficat, “the effect of a core biopsy is similar to that of a decompression operation for a nerve tunnel syndrome or a fascial release for a muscle compartment syndrome.”

Surgical Technique

With the patient supine on the fracture table, the hip is placed in a neutral position. A 5–10° internal rotation is helpful, as it places the femoral neck parallel to the floor. A 8–10 cm. skin incision is made, starting proximal to the vastus lateralis tubercle of the greater trochanter, and extending distally along the posterior edge of the vastus lateralis itself.

The fascia lata is therefore split corresponding to the skin incision. The vastus lateralis is stripped anteriorly with a periosteal elevator in order to expose the lateral aspect of the greater trochanter. A cortical window is made just below the trochanteric ridge, wide enough to allow the introduction of the trephine (Fig. 2).

Fig. 2
figure 2

Entry point for the introduction of the trephine

Full size image

A hollow trephine, 6–8 mm. in diameter, is then introduced into the neck of the femur through this window. The coring device is directed into the anterosuperior portion of the femoral head until the tip arrives within 5 mm. of the subchondral plate (Fig. 3). Its manipulation is controlled, both in the anteroposterior and lateral planes, with an image intensifier. A second channel is made with a smaller trephine in a different direction. As hyperpressure exists in the bone marrow around the necrotic area, some authors recommend to drill up to, but not through, the border of the necrosis.

Fig. 3
figure 3

A hollow trephine, 8 or 6 mm. in diameter, is introduced into the neck of the femur through this window: the coring device is directed into the anterosuperior portion of the femoral head

Full size image

The core specimens thus obtained are then sent for histological examination. The forage channels are left open.

There are two reported methods of core decompression: large diameter trephines and small diameter drills.

The most common method, the 8–10 mm. trephine, is completed under fluoroscopy with the core track either being left open or filled in with bone graft.

After-Care

Aftercare is strongly dependent on the patient's age, weight and collaboration.

Usually, partial weight-bearing is not indicated before 40 days. Full weight-bearing is permitted at 3 months.

Discussion

The efficacy of core decompression is controversial, and the Orthopaedic literature has not clearly defined the role for this procedure, considered as the only treatment [16, 20].

Ficat reports good clinical results in 94 % of Stage I and 82 % of Stage II of his classification. Fairbank reported the long-term results of core decompression, with a 15-year survival rate of 90 % in Stage I, 66 % in Stage II, 23 % in Stage III [9].

These results are similar to those reported by other authors: Smith, with an average 3 year follow-up, reported the failure of the core decompression (defined as performance of a subsequent operation) in 16 % of Stage I, 80 % of Stage II, and 10 % of Stage III hips [37].

It seems reasonable to conclude that the rate of success of the core decompression is largely dependent on:

  1. 1.

    The accuracy of the diagnosis: excluding transient osteoporosis, reflex sympathetic dystrophy;

  2. 2.

    The correct timing: core decompression gives excellent results in Stage I and some Stage II patients, while it is unacceptable for later stages [12];

  3. 3.

    The correct surgical procedure: neck fracture has been described as a frequent complication of core decompression [15]. A simple means to eliminate this complication is to avoid an approach which is too distal.

Recent Innovations

Hormone-Enhanced Bone Grafts

The most recent surgical innovations currently under investigation represent modifications of standard core decompression. There is interest in the development of biologically-based therapies that can enhance core decompression with either osteoinductive (bone morphogenic protein) or osteogenic (mesenchymal cells) agents that have the clinical potential to provide better results for larger lesions [31].

A pair of retrospective studies has evaluated the efficacy of bone morphogenetic protein-enhanced bone graft in preventing disease progression in osteonecrosis of the femoral head. Mont et al. [28] used a modified trap-door technique and bone morphogenetic protein–enriched bone graft substitute through a window at the femoral head-neck junction in 23 patients and reported successful clinical results in 18 of 21 hips at a minimum follow up of 36 months.

Mesenchymal Stem Cells

The use of bone marrow-derived mesenchymal stem cells (MSCs) as an adjunct to core decompression has shown promising results in one large series, and statistically significant improvements in outcomes in a small pilot randomized control trial. In vitro and in vivo data suggest that MSCs are capable of undergoing osteogenic differentiation and mediating bone mineralization. Additionally, their role in improving the haemodynamic function of injured myocardium in animal models has been attributed to the elaboration of paracrine factors such as vascular endothelial growth factor that lead to neovascularization. Finally, it is possible that the regenerative potential of MSCs is influenced by remodelling of the extracellular matrix, an important factor in reversing the pathologic process that initiates osteonecrosis [17].

Metallic Porous Implants

The use of adjuncts such as porous tantalum implants in combination with core decompression has the potential to provide the structural advantages of bone graft without the associated risk of autograft harvest or the infectious complications associated with allograft bone.

Porous tantalum is indicated mainly for the manegement of early stage (Stage I or II) osteonecrosis of the femoral head in patients who do not have chronic disease [21].

The use of a tantalum implant has been reported in two studies. Tantalum is a light metal that has a high yield to stress. In these studies, porous tantalium rods were used to potentially allow bone growth to occur while providing support. While the short-term results in these studies compared favourably to other core decompression techniques, longer follow-up is needed to more fully assess the efficacy of this procedure [24].

Procedures Aimed at Moving the Necrotic Area Out of the Weight- Bearing Zone

Many different osteotomy techniques have been proposed to move the necrotic area, which is usually located in the anterosuperior part of the femoral head, out of the weight-bearing zone. The type of osteotomy he performed is controversial, as osteotomies have been proposed on every plane, each with its own rationale.

Merle d’Aubigné first introduced varus osteotomy to load the intact far-lateral part of the femoral head [25]; Saito gave as an indication for varus osteotomy the presence of more than one-third of the head intact, laterally [33].

A valgus-extension osteotomy has been proposed by Pauwels and Bombelli, in order to move the necrotic area laterally and posteriorly, and to enlarge the weight- bearing surface by loading the so-called “capital drop” osteophyte [3, 22, 30]. In opposition to this technique, it has been remarked that this osteophyte is poorly developed in Stage III, contrary to Stage IV necrosis, in the presence of osteoarthritic modifications that make this indication debatable for a salvage procedure [8, 34].

As the anterosuperior part of the femoral head is usually involved in the necrosis, an osteotomy aimed at moving the necrotic area anteriorly seems to be the most logical solution. This may be achieved either by rotating the head and neck anteriorly, as proposed by Sugioka, or by performing an anterior wedge-flexion osteotomy, as proposed by Schneider [12, 13, 39].

Sugioka, by using the Japanese roentgeno- graphic staging system of idiopathic femoral head necrosis (comparable to the Ficat classification) reported a 73 % success rate of his procedure in Stage III and a 70 % rate in Stage IV, after a follow-up from 3 to 16 years [40]. However, the rotational osteotomy did not give the same good results outside Japan: Tooke had a 41 %, rate of revision to hip replacement at 39 months [41]. Dean, reporting a 83 % rate of early failure, concluded that the “Sugioka anterior rotational osteotomy … should not be performed in Caucasian patients. It does give satisfactory early pain relief, perhaps through joint denervation produced by the capsulotomy, but the procedure appears to compromise further the blood supply of the femoral head…” [7]. He gave as an explanation for these results the fact that the anterior rotation of the femoral head stretches the quadratus femoris, therefore flattening and obstructing the posterior branch of the medial circumflex artery. This author hypothesises that the posterior capsule of the hip may be more lax in Japanese patients, allowing for anterior rotation of the neck without compromising the flow through this artery.

Schneider, in 1969, at the Congress of the Orthopaedic Swiss Society, proposed a flexion intertrochanteric osteotomy aimed at rotating the intact posterior portion of the femoral head into the weight-bearing portion of the acetabulum. This osteotomy became popular in Europe due to its efficacy when carefully planned and performed on well-selected patients [35].

Technique of Flexion Osteotomy

Planning the Osteotomy

Pre-operative planning of the osteotomy is based on evaluation of the following parameters:

  1. 1.

    Radiological: an A.P. view to evaluate the angle of the neck (valgus or varus), the morphology of the trochanteric area, the condition of the opposite hip and the likelihood of its future deterioration. These criteria enable the surgeon to select the right point to introduce the osteotome, the level of the osteotomy, and the amount of displacement. Two cranio-caudal views at 15° and 30° to assess the amount of flexion are necessary. An axial view to evaluate the extent of the healthy posterior sector of the head and any anteversion of the neck. MRI, even if absolutely mandatory in the diagnosis of the osteonecrosis, gives no supplementary information useful for the osteotomy planning. A CT scan allows a precise evaluation of the location and the extent of the lesion.

  2. 2.

    Amount of flexion: it is necessary to relieve the necrotic area of load when the hip is flexed about 20°, as occurs during walking.

    Therefore, it is necessary to add a further 20° to the flexion which makes the necrotic area disappear in the cranio-caudal x-ray. The necessary angle of flexion (α) is calculated. In recent years, we have tended to increase the limit, because it is not always possible to achieve the exact pre-operative calculation of correction. However, we suggest that the limit should be 50° of flexion.

  3. 3.

    Fixation device: the osteotomy site is stabilized with an AO right-angle or 95° blade-plate; usually, the blade does not exceed 50 mm. in length.

Surgical Procedure

With the patient supine on the fracture table, the leg is draped to allow free movement during the operation [12, 13, 35].

The anterolateral approach is used as a rule: the anterior portion of the gluteus medius and the tendon of the gluteus minimus are cut, and longitudinal capsulotomy is performed to allow inspection of the femoral head, eliminating the need for intra-operative radiography. Some surgeons do not perform the arthrotomy, in which case each step of the operation requires fluoroscopic control. Guide-wires are inserted to achieve the planned flexion. The first osteotomy “a–b” is performed at a right angle to the diaphysis and just proximal to the lesser trochanter, and is not completed (Fig. 4). The second osteotomy is performed on the “c–d” plane according to the α angle calculated pre-operatively. In order to achieve a complete congruence of the osteotomy surfaces, “a–b” should be equal to “c–d.” This can be achieved either directly or by subsequent trial osteotomy (d’, d”).

Fig. 4
figure 4

A–B = first osteotomy; C–D = second osteotomy; E = distance between the second osteotomy and the entry point of the blade

Full size image

The seating chisel for the blade-plate has to be inserted parallel to the plane of the second osteotomy “c–d,” inferiorly in the femoral head and posteriorly in the neck. Its entry point is 1.5–2 cm. proximal to the “c–d,” and should correspond to the middle portion of the “c–d” length (Fig. 5).

Fig. 5
figure 5

The operative field: the seating chisel for the blade-plate has to be inserted parallel to the plane of the second osteotomy “C–D”; its entry point has to be 1.5–2 cm. proximal to the “C–D,” and should correspond to the middle portion of the “C–D” length

Full size image

The blade is inserted almost completely, the first osteotomy is completed and the bone wedge removed to allow contact between the osteotomy planes. Before removing the wedge, the osteotomy surfaces are marked to avoid rotational errors. The plate is positioned on the femoral shaft, the osteotomy site is compressed and the plate secured with a single screw. Rotation is checked, and, if satisfactory, the plate is fully secured: if not, the holding screw is released and the necessary rotational adjustment made before securing the plate (Fig. 6).

Fig. 6
figure 6

Pre-operative and post-operative X-rays of a flexion osteotomy; the osteotomy site is stabilised with an AO right-angle or 95° blade-plate. The blade usually does not exceed 50 mm. in length

Full size image

If the α angle, i.e., the flexion to be reached, exceeds 30°, psoas release is mandatory to avoid flexion deformity.

Some authors have combined the osteotomy with curettage of the necrotic bone and packing of the remaining cavity with autogenous bone graft from the iliac crest [34].

Errors and Pitfalls

  1. 1.

    Placing the blade too anteriorly, thus jeopardising the congruence of the osteotomised surfaces [13];

  2. 2.

    Inserting the blade too distally: the amount of bone between the blade and the osteotomy is reduced and consequently the stability of the fixation is threatened;

  3. 3.

    Placing the blade correctly from the point of view of the osteotomy, but not in the axis of the femoral neck. The main risk is that the blade may penetrate the posterior cortex, particularly in anteverted hips, with possible damage to the blood supply of the head.

  4. 4.

    Omitting to tenotomise the ileo-psoas tendon and Bertin's ligament when high degree flexion osteotomies are performed; the risk in such circumstances is a fixed flexion deformity.

After-Care

A regime of passive mobilisation of the hip is begun in the first few days: active abduction and flexion are usually contra-indicated before 40 days. If there is a tendency to hip flexion, patients are told to lie prone for a short period every day during the first few months; weight-bearing is not allowed before 3–6 months.

Discussion

The results of the series reported in the literature are difficult to compare, due to the differences in the indications for this operation. Most authors agree that intertrochanteric osteotomy for the treatment of AVN of the hip does not give totally predictable results, and that patient characteristics, pathogenesis of the AVN, stage and extent of the lesion, as well as pre-operative range of motion, all play an important role in the final outcome [12, 14, 22, 34].

  1. 1.

    Patient characteristics: young patients are the best candidates for intertrochanteric osteotomy. The upper age limit is generally around 45 years. Furthermore, the operation must be avoided if there is poor patient compliance or if he/she is not highly motivated to accept a lengthy period of disability.

  2. 2.

    Pathogenesis: “idiopathic” AN represents the best indication, whilst intertrochanteric osteotomy should he avoided in cases of underlying metabolic disease or in the presence of a systemic condition that has been treated with chemotherapy or corticosteroids.

  3. 3.

    Stage and extent of the lesion: the Ficat Stage III lesions or late Stage II can be successfully treated with intertrochanteric osteotomy, while Stage IV is a contra-indication. As to the extent of the necrotic area, the best results can be obtained in those patients who have a necrosis angle of less than 200°, calculated according to Kerboull as the sum of the arc of the surface involved measured in A-P and lateral radiographs [5, 14, 19]. We give a more limited indication, avoiding osteotomy if the lesion has reached the surface in the cranio-caudal views at 30°, or if the lesion exceeds the fovea shape in the A–P view.

Range of motion: all authors agree that the passive range of motion has to be normal or slightly reduced: flexion should be to at least 110°, and adduction to 20° [14].

Scher reported 87 % satisfactory outcomes at 10 years after intertrochanteric flexion osteotomy [34]. This result is in agreement with our findings and has led us to justify this operation with these very limited indications, as well as on the basis of the absence of serious technical problems in case of a subsequent total hip prosthesis.

Procedures Aimed at Restoring Bone Blood Supply in the Necrotic Area

Judet [18] first introduced the muscle-bone pedicle graft of the greater trochanter for fractures of the femoral neck. Meyers and Palazzi used a similar procedure in treating Ficat’s Stages I-II AVN, reporting good results. Implantation of a vascular pedicle into the femoral head was attempted by Hori, who reported new bone formation around the vascular pedicle [47]. None of these procedures could provide a mechanical support to the subcondral bone. Several papers report good results with the use of free vascularised fibular grafting in treating AVN of the femoral head.Vascularised fibular grafting introduces not only a source of mesenchymal stem cells and a vascular supply, but also a structural bone graft able to provide articular support [36, 44, 46].

Non-Vascularised Graft

Non-vascularised bone grafting is more invasive than core decompression and has typically been used in hips with a femoral head collapse of <2 mm. or in hip in which core decompression has been unsuccessful.

Core decompression and non-vascularized bone-grafting techniques are viable options to avoid the need for additional surgery in patients with early stage osteonecrosis of the hip [24].

Vascularised Graft

The presence of a symptomatic Stage II to Stage IV lesion in a patient younger than 50 years of age is the main indication for fibular grafting. Narrowing of the joint space and acetabular involvement are contra-indications to surgery which is designed to preserve the femoral head. Collapse of the femoral head and depression of more than 2 mm. are relative contra-indications for a free vascularized fibular graft.

The fibula, with its large peroneal vascular pedicle, nutrient artery and periosteal blood supply, became a favoured source of vascularised bone.

It is now generally agreed that osteonecrosis of the femoral head can be arrested in most patients if the procedure is performed before the developement of a subchondral fracture (Stage I and II, Steinberg classification). The results are more unpredictable when there is collapse of the femoral head. In Stage III, fibular grafts may arrest or delay the progression of desease [21].

Urbaniak et al. [43] reported a probability of conversion to THR within 5 years after primary procedure of 11 % for Stage II hips, 23 % for the Stage III hips, 29 % for the Stage IV hips and 27 % for the Stage V hips. They described an 83 % survival of the fibular graft in 646 procedures after follow-up ranging from 1 to 17 years.

Plakseychuk et al. [32], in attemp to compare vascularised with non-vascularised fibular grafting, presented 220 hips treated by vascularised grafting with a minimum follow-up of 3 years. In patients with pre-collapse Stage I and II, the rate of survival of the femoral head at 7 years was 86 %. Similar results were observed in a review of 1,303 vascularised fibular grafts performed at seven centres with a minimum follow-up of 2 years [6].

The success of the procedure may be due to:

  • Decompression of the femoral head, which may halt the ischaemia caused by increased intraosseous pressure;

  • Excision of the necrotic bone beneath the weight-bearing region that might inhibit revascularisation of the femoral head;

  • Buttressing of the articular surface with the vascularised graft by primary callus formation, augmented by cancellous bone graft, which has osteoinductive and osteoconductive factors;

  • Protection of the healing construct by a period of limited weight-bearing.

The factors influencing the outcome are:

  • Aetiology of the necrosis,

  • Stage of the disease,

  • Size of the necrotic segment.

According to Berend et al. [2] the amount of flattening or pre-operative linear collapse does not significantly affect survival or functional outcome.

Marciniak et al. [23] suggested that there is no relationship between the initial radiographic stage of the disease and the functional outcome or the rate of graft survival.

Although success or failure of the free vascularised fibular graft procedure has been more closely correlated with the size of the lesion and the amount of collapse of the femoral head, the patient’s demographic factors should also be considered when formulating a treatment plan. Age, general health status activity level, the range of movement of the hip and co-morbidities appear to influence the functional outcome.

Technique of Vascularised Fibular Graftimg

Pre-Operative Planning

Besides the standard pre-operative imaging some authors suggest the need for arteriography to confirm that the patient's vascular pattern, distal to the trifurcation of the femoral artery, is intact.

Surgical Procedure

The operation can be performed either with the patient supine or in lateral decubitus on the fracture table [4, 43, 44]. The hip and the lower limb are prepared and draped as a single sterile field; a tourniquet is placed on the thigh. Usually, the femur and the fibula are approached simultaneously by two separate teams to lower operative times.

Operative Procedure on the Femur

The femur is approached through an anterolateral incision: the lateral aspect of the proximal part of the femur is exposed through the interval between the tensor fasciae latae and the gluteus medius (Fig. 7). The vastus lateralis is then reflected from the vastus ridge for approximately 5 cm. to expose the bone (Fig. 8).

Fig. 7
figure 7

The fibular graft is inserted into the femoral head beneath the subchondral bone, within the cancellous bone graft: the pedicle is located superiorly and anteriorly. The fibular graft is then secured to the femur with a Kirschner wire

Full size image
Fig. 8
figure 8

Vascularized fibular grafting: post-operative control X-rays

Full size image

Anteriorly, the origin of the vastus intermedialis is carefully released to create a gap that provides a shorter route for the ascending vessels. To avoid damaging of vessels and the femoral nerve, the dissection must not exceed the fat layer medial to the vastus intermedialis. The donor vessel is usually the ascending branch of the lateral circumflex artery, which can be identified as it runs laterally between the vastus intermedialis and the rectus femoris. The artery and two veins are carefully mobilised for a length of about 4 cm; haemostatic clips are placed on the end of each of the three vessels (Fig. 9).

Fig. 9
figure 9

Haemostatic clips placed on the end of each of the three vessels

Full size image

Under fluoroscopic control, a guide-wire is inserted into the femoral neck and directed into the centre of the necrotic area. Cannulated reamers are progressively used over the guide-pin; the maximum size of the reamer depends on the largest diameter of the fibular graft, being usually 16 mm. in females and 19 mm. in males. The reaming extends to within 3–5 mm. from the articular surface of the femoral head. Cancellous bone graft is then harvested from the greater trochanter and packed into the cavity.

Operative Procedure on the Fibula

The tourniquet is inflated to an appropriate pressure.

The fibula is approached through a lateral incision. The muscles are reflected off the fibula. The deep fascial layer is incised anteriorly and posteriorly. Anteriorly, the interosseous membrane is incised along its length. The pedicle lies posteriorly, directly under the flexor hallucis longus.

The distal cut is performed at least 10 cm. from the tip of the lateral malleolus, while the second cut is usually 15 cm. proximal to the first. In the distal wound, the pedicle is dissected free from the muscle for a length of 5 cm. Two haemostatic clips are placed across the pedicle. After the pedicle has been transected, the fibular graft is freed from the leg.

40 ml. of heparinised lactate Ringer’s solution are injected into the artery and both veins. A 3-0 absorbable suture is passed circumferentially around the distal part of the periosteum and vascular pedicle, to avoid stripping of the periosteum and the pedicle at the insertion of the graft into the core.

Placement of the Graft

The fibular graft is inserted into the femoral head beneath the subchondral bone, within the cancellous bone graft: the pedicle is located superiorly and anteriorly. The fibular graft is then secured to the femur with a Kirschner wire (Fig. 10). With the help of an operating microscope, the arterial and venous anastomoses are performed with 8-0 or 9-0 interrupted nylon sutures. Both incisions are closed over drains.

Fig. 10
figure 10

Fibular graft isecured to the femur with a Kirschner wire

Full size image

After-Care

Non-weight-bearing ambulation is begun with a walker or crutches as of the first post-operative days, and continued for 6 weeks, after which partial weight-bearing is started, progressing to full weight-bearing at 6 months after surgery.

Discussion

According to Brunelli and Urbaniak, this technique is indicated in symptomatic AVN developed to Ficat Stage II or III in relatively young patients (under 40 years according to Urbaniak, under 55 according to Brunelli), and who are well-motivated. These indications appear to be very close to those given for the rotational osteotomy; however, we did not find any study which compared the outcome of these two techniques.

Recently, Scully published a comparison of the outcome of core decompression and vascularised fibular grafting. By using total hip arthroplasty as the end-point, the survival rate was significantly higher at 50 months for vascularised fibular grafting, both for Ficat Stage II (89 %) and Stage III (81 %) lesions [36]. Even in the presence of such promising results, it has been pointed out that vascularised fibular grafting requires a great deal of technical expertise, implies a prolonged period of restricted weight-bearing and is affected by considerable morbidity associated with the graft donor site [26, 45].

Conclusions

The place remaining for conservative surgery of avascular necrosis is becoming more and more reduced, due to progress in the field of total hip replacement and to the increasing demand from patients for an immediate and optimal result. The majority of the surgical procedures described here are more technically- demanding than a hip replacement, result in a lengthy period of disability and do not give totally predictable results. Nonetheless, conservative surgery still has an important place in carefully-selected patients and in the early stages of the disease.

Researchers in many Orthopaedic fields have recently focussed their attention on the effects of growth and differentiating factors on the bone healing process: cytokines, bone morphogenetic proteins and angiogenic factors have been used successfully in animal experimental models of AVN. Their use in human therapy, in combination with conservative surgical techniques, as suggested by Mont, may represent the next, fundamental step in the treatment of this crippling disease [27].