A current review of non-vascularized bone grafting in osteonecrosis of the femoral head

Abstract

Over the past three decades, non-vascularized bone grafts have been demonstrated to be viable treatments for pre- and early post-collapse osteonecrosis of the femoral head; however, there are limited reviews on this topic. Therefore, the purposes of this review are to (1) provide a summary of the different surgical techniques and their respective clinical outcomes and (2) evaluate new adjunct therapies. Originally, non-vascularized bone grafting was performed using the Phemister technique with varying results. More recently, newer techniques such as the lightbulb and trapdoor are used to place non-vascularized bone grafts with excellent results. The addition of various biological agents has demonstrated positive results; however, further studies are needed to confirm the best appropriate indications and to elucidate long-term results.

Introduction

Over the past three decades, non-vascularized bone grafting (NVBG) has demonstrated to be a viable treatment option for patients who have osteonecrosis of the femoral head (ONFH). These techniques are indicated for patients who have pre-collapse (Ficat stages I and II) [1–3] or early post-collapse disease (Ficat stage III), where the articular surface is intact or with less than 2 mm of depression of the femoral head [1, 4].

In 1949, D.B. Phemister [5] was one of the first authors to describe the use of a non-vascularized bone graft (NVBG) from the tibia for the treatment of ONFH. Since that time, there have been newer techniques investigated for the introduction of NVBG that have become more widely used, including the lightbulb and trapdoor methods [6–8]. The theory behind this technique is the debridement of weak necrotic bone down to good viable bleeding bone, as well as providing a scaffold for the formation of new bone, thus providing stability to the femoral head [1, 4].

Currently, there is an abundance of studies regarding the use of various bone grafts and their outcomes. However, given the widespread prevalence of ON, it is necessary to provide an in-depth review on the appropriate use of NVBG in these patients. Therefore, the purposes of this review are to (1) provide a brief summary of the different surgical techniques, (2) report their clinical outcomes, and (3) evaluate new adjunct therapies for NVBG.

Surgical techniques

When examining outcomes, it is important to consider the role of surgical technique (Table 1). We will examine the use of the Phemister, lightbulb, and trapdoor techniques and their outcomes.

Table 1.

Various studies on NVBG outcomes since 2001

Author (year)	Number of hips	Surgical technique	Stage of disease	Mean follow-up (months)	Hip survivorship (%)
Windisch et al. (2014) [29]	18	Phemister	ARCO II and III	NR (12 to –)	78
Gagala et al. (2013) [30]	13	Trapdoor	ARCO II to IV	33 (18 to 75)	62
Zhang et al. (2012) [31•]	85	Lightbulb	ARCO I to III	NR (24 to –)	85
Wang et al. (2012) [32••]	28	NR	ARCO I to III	104 (93 to 108)	56
Wei and Ge (2011) [33]	223	Phemister	ARCO II to III	24 (7 to 42)	81
Hsu et al. (2011) [14]	62	Phemister	Steinberg I and II	46 (24 to 107)	63
Tetik et al. (2011) [34]	15	Phemister	Ficat II and III	42 (16 to 114)	87
Yang et al. (2010) [35]	54	Phemister	Steinberg I to III	NR (36 to 78)	84
Wang et al. (2010) [36]	138	Lightbulb	ARCO II to III	25.4 (7 to 42)	68
Yuhan et al. (2009) [37]	11	Lightbulb	ARCO II to III	61 (30 to 103)	73
Seyler et al. (2008) [21]	39	Trapdoor	Ficat II to III	36 (24 to 50)	67
Keizer et al. (2006) [15]	80	Phemister	Ficat I to IV	84 (– to –)	56
Israelite et al. (2005) [38]	276	Phemister	Steinberg I to IV	NR (24 to 145)	62
Mont et al., (2003)a [18]	21	Lightbulb	Ficat II and III	48 (36 to 55)	86
Rijnen et al. (2003) [16]	28	Phemister	ARCO II to IV	42 (24 to 119)	71
Steinberg et al. (2001) [39]	312	Phemister	Steinberg I to IV	48 (24 to 155)	64

NR and – not reported, ARCO Association of Research Circulation Osseous

^aBone graft enriched with bone morphogenic protein (BMP)

Phemister technique

The Phemister technique, also referred to as cortical strut grafting, was first described in 1949 by Phemister [5] and popularized by Bonfiglio and Bardenstein [9] a decade later. It involves removing a 7- to 9-mm-diameter cylindrical core from the femoral head and neck (similar to core decompression), which is then replaced by a cortical graft removed from the tibia, fibula, or ilium. In 1970, Boettcher et al. [10] evaluated the clinical and radiographic outcomes of 38 hips with ONFH treated with cortical strut grafting. The authors defined a satisfactory outcome as minimal functional restriction, moderate, minimal, or no pain and a score on the Iowa hip-rating scale of at least 70 points. After a mean follow-up of approximately 6 years (range, 2 to 17 years), they found satisfactory results in 79 % of hips. However, this same cohort was evaluated by Smith et al. [11] after a mean follow-up of 14 years (range, 4 to 27 years), and they noted that of those not deceased or lost to follow-up, only 16 hips (57 %) had satisfactory outcomes.

In 1991, Buckley et al. [12] evaluated a cohort of patients in the pre-collapse stages of ON treated with NVBG (n = 20 hips). After a mean follow-up of 8 years (range, 2 to 19 years), only two hips (10 %) progressed to require a total hip arthroplasty (THA). Similarly, Nelson and Clark [13] evaluated the disease progression in a cohort of patients that underwent grafting using the Phemister technique (n = 52 procedures), with a minimum follow-up of 24 months (range, 24 to 144 months). Of the 17 hips grafted at stage 2, 13 (76 %) progressed at least one stage. Of the 11 hips grafted at stage 3, 7 of them progressed. Therefore, the authors concluded that this grafting method is not recommended once excessive femoral head collapse is present.

When evaluating candidates for the Phemister technique, some have studied whether or not lesion size or symptoms can impact outcomes. Hsu et al. [14] assessed the outcomes of this technique in those with asymptomatic (n = 31 hips) and symptomatic small osteonecrotic lesions (n = 31 hips). After a mean follow-up of approximately 46 months (range, 24 to 107 months), 10 of the 31 asymptomatic hips (32 %) and 13 of the 31 symptomatic hips were eventually converted to THA (p = 0.3). Furthermore, the interval time to conversion to THA was not significantly different between the two cohorts (15 versus 13 months, respectively; p = 0.18).

More recently, the outcomes of this technique have been less favorable, as this does not adequately debride the lesion. Keizer et al. [15] evaluated the use of autogenous cancellous bone grafting through a core tract in a cohort of patients who had Ficat stages 1 to 4 ONFH (n = 78 hips). After a mean follow-up of 7 years, 34 hips (44 %) required further procedures. Furthermore, they found a mean 5-year clinical survivorship of 59 %, which decreased to 44 % at 10-year follow-up.

In addition, this technique may work best in a certain subset of patients. Rijnen et al. [16] assessed the outcomes of NVBG in conjunction with core decompression in a consecutive series of patients (n = 28 hips). After a mean follow-up of 42 months (range, 24 to 119 months), 64 % (18 hips) of the hips had clinically successful outcomes as defined by minimal or no pain and 50 % (14 hips) showed no radiological signs of disease progression. However, 29 % (eight hips) underwent THA by their latest follow-up. The authors noted that those with the best outcomes were younger than 30 years of age while those with the worst outcomes had ON associated with corticosteroid use. Furthermore, the authors concluded that this salvage procedure was a viable option, as it can delay THA and does not interfere with the outcomes or complexity of THA, if eventually required.

In summary, bone grafting using the Phemister technique can be effective. However, because some studies have reported unfavorable results, the majority of surgeons have investigated other techniques of bone grafting.

Lightbulb technique

Ganz and Buchler [6] first introduced the method of using the femoral neck as a conduit for introducing bone graft. After performing an osteotomy, the authors described filling the femoral head defect with cancellous bone graft. However, this procedure has evolved to be performed through a cortical window in the femoral neck, rather than in conjunction with an osteotomy. This technique involves the formation of an anterior hip arthrotomy via a Watson-Jones or Smith-Petersen approach [1]. Then, a 2 × 2 cortical window is made at the femoral head neck junction. The cortical window IS then removed, allowing for access to the femoral head and the removal of any necrotic bone with a high-speed burr [1]. Following necrotic segment debridement, the lesion is visualized with a light stick to ensure adequate lesion removal (presence of bleeding bone all around lesion cavity). The lesion can be packed with various forms of bone grafting and augments, and the cortical window is then replaced and secured with two or three divergent bioabsorbable Steinman Pins.

The “lightbulb” technique was popularized by Rosenwasser et al. [7], who assessed the outcomes of using this method in a cohort of patients who had Ficat stage II or III disease (n = 15 hips). After a mean follow-up of 12 years (range, 10 to 15 years), 13 hips remained symptom free without any signs of disease progression. Hence, the authors concluded that this technique of grafting may delay or prevent the need for THA. Using the same technique, Itoman and Yamamoto [17] evaluated the use of cortical iliac bone graft in patients who had Ficat stage II or III ONFH (n = 38 hips). After a mean follow-up of 9 years (range, 2 to 15 years), with good outcomes being defined as requiring no further surgery and being pain free, the authors reported good clinical outcomes in 23 (61 %) of the hips.

Similarly, Mont et al. [18] evaluated the outcomes of 21 patients who had early stage ONFH (Ficat II or III) that underwent NVBG using the lightbulb technique. After a mean follow-up of 48 months (range, 36 to 55 months), Haris Hip Score (HHS) improved from a pre-operative mean of 48 (range, 25 to 62 points) to 91 points (range, 80 to 100 points) at final follow-up. In addition, 86 % of the hips (n = 18 hips) had successful clinical and radiographic outcomes. Therefore, the authors concluded that this procedure could be indicated in patients with pre-collapsed lesions with less than 2 mm of head depression or hips that have failed core decompression alone.

In summary, this technique, which uses the femoral neck as a conduit for the insertion of bone graft, has shown positive outcomes. Even still, there is one other technique that has attempted to further improve upon these positive outcomes following NVBG.

Trapdoor technique

This technique is nearly identical to the approach of the lightbulb technique; however, rather than removing a cortical window from the femoral head-neck junction, the femoral head is dislocated and a window is created through the collapsed articular cartilage. A 2 × 2-cm flap is created from the chondral surface, allowing for necrotic bone removal with a high-speed burr. As per the last technique, the lesion is filled with bone graft, closed, and secured with bioabsorbable pins.

This technique was first reported in 1983 by Meyers et al. [8]. They reported marked improvements in pain and function in 15 out of 21 hips (71 %) after a minimum follow-up of 18 months (range, 18 to 60 months). About a decade later, this method of bone grafting was implemented by Ko et al. [19] in a cohort of teenagers who had ONFH (n = 14 hips). After a mean follow-up of 53 months (range, 24 to 108 months), none of the hips required further procedures at their latest follow-up. Thus, the authors concluded that trapdoor bone grafting may delay the need for fusion in teenagers with ON.

Similarly, Mont et al. [20] assessed the use of this technique in 30 patients who had Ficat stage III or IV ONFH. After a mean follow-up of 56 months (range, 30 to 60 months), the HHS improved from a mean of 41 points (range, 31 to 64 points) pre-operatively to 92 points (range, 80 to 100 points) at latest follow-up. Furthermore, 20 of 24 stage III hips (83 %) had good or excellent radiographic outcomes; however, four of the six stage IV hips required THA. As such, the authors concluded that this technique of introducing bone graft may be successful in those with Ficat stage III ON; however, once the disease has progressed to end-stage ON, outcomes are less favorable. More recently, Seyler et al. [21] evaluated the use of this technique in 39 hips with Ficat stage II and III ONFH. After a mean follow-up of 36 months (range, 24 to 50 months), the mean HHS significantly increased from 50 points (range, 28 to 76 points) pre-operatively to 75 points (range, 27 to 100 points) post-operatively (p < 0.001). Furthermore, 26 (67 %) hips did not require THA at latest follow-up. Of those with Ficat stage II disease (n = 22 hips), 82 % (n = 18 hips) did not require THA. The authors concluded that NVBG using the trapdoor technique may delay the natural progression of ON, with the most success in Ficat stage II ON.

However, as with other techniques, those in more advanced stages of ON may not benefit from this procedure. Although there is a paucity of studies critiquing and analyzing the trapdoor technique and its outcomes, it has been reported that it may prevent or delay THA in patients with ON. Additionally, unlike the Phemister and lightbulb techniques, this method of introducing bone graft may benefit those in the early post-collapse stages of the disease.

Adjuncts to NVBG

Despite the positive results that have been reported with the use of NVBG, osteonecrosis is often a progressive disease, and to improve outcome, orthopedists have attempted to further improve upon these techniques with the addition of several adjunctive methods such as bone morphogenetic proteins and stem cells.

Use of BMPs

Bone morphogenetic proteins have been shown to stimulate new bone formation, initiate angiogenesis, and promote articular cartilage repair [22, 23]. Lieberman et al. [24] evaluated the use of bone morphogenetic protein (BMP) in conjunction with NVBG in a series of 17 patients who had Ficat stage II and III ONFH. After a mean follow-up of 53 months (range, 26 to 94 months), 14 hips had pain relief and no radiographic progression of disease. In addition, Seyler et al. [21] augmented their bone grafts with various BMP subtypes and found that no patients (n = 39) required further procedures at latest follow-up. Therefore, the authors concluded that not only was the trapdoor technique of grafting acceptable, but also the various growth and differentiation factors that were infused in their grafts may have also assisted in improving outcomes.

More recently, Papanagiotou et al. [25•] assessed the effect of fibular NVBG with recombinant human BMP-7 (rhBMP-7) on clinical and radiographic outcomes for the treatment of seven hips with Steinberg stage II and III ONFH. After a 2-year follow-up, there was a marked increase in mean HHS from 34 points (range, 22 to 50 points) pre-operatively to 87 points (range, 70 to 90 points) post-operatively (p > 0.05). Similarly, a randomized trial by Sun et al. [26] evaluated the use of rhBMP-2 with impact NVBG (n = 43) compared to NVBG alone (n = 36) for the treatment of pre-collapse ONFH. After a mean follow-up of 6 years (range, 5 to 8 years), the authors found a marked difference in survival rates between the control and experimental cohorts (72 versus 82 %, respectively). Furthermore, there was a slight difference in mean post-operative HHS between the two cohorts (78 versus 82 points, respectively).

These positive results suggest that the use of BMP in conjunction with NVBG may improve outcomes in patients with ONFH. Larger, prospective randomized studies are needed to determine the usefulness of using BMP with NVBG as routine therapy.

Stem cells

Most recently, there has been exploration into the use of stem cells as an adjunct to NVBG, and positive results have been seen in small case series. Wang et al. [27] evaluated the efficacy of such a method using bone marrow mononuclear cells as an adjunct to NVBG. After a mean follow-up of 24 months (range, 9 to 36 months), there was a 100 % hip survival rate and the mean Harris Hip Score improved from 64 to 85 points. Similarly, Aarvold et al. [28] assessed the use of skeletal stem cells (SSC) in conjunction with NVBG in a series of patients (n = 5 hips). At the 44-month follow-up appointment, three hips remained asymptomatic. However, the other two hips were converted to THA due to collapsed femoral heads. In an in vitro analysis of the bone graft with SSC in the patient converted to THA, there were signs of viability and proliferation.

Conclusion

The use of NVBG is an effective hip-preserving operation that may be considered in a selected group of patients who have ONFH. Currently, the most widely agreed upon indications for this treatment are in the pre-collapse stages of ON [1–3] or the early post-collapse stages where there is less than 2 mm of depression of the femoral head and the articular surface is intact [1, 3]. The use of various surgical techniques is a matter of surgeon preference and is an area of active investigation. In addition, future research should focus on the appropriate biological adjuncts to use with NVBG and comparisons of outcomes with other hip-preserving surgeries.