Biomechanical analysis of iliac crest loading following cortico-cancellous bone harvesting

Iliac crest bone harvesting is a frequently performed surgical procedure that is widely used to treat bone defects in orthopedic and trauma surgery as well as in reconstructive surgery and oral and maxillofacial surgery. Harvesting at the anterior iliac crest provides grafts with all properties required. Advantages such as fair bone quantity (cancellous bone, bicortical graft, tricortical graft, vascularized graft) when structural support is needed combined with the outstanding feature of autologous bone containing progenitor cells as well as growth factors make this relatively easy procedure to be considered as the “gold standard” [1, 2]. Nonetheless, post procedural complications such as donor site pain, gait disturbance, numbness, and fractures of the iliac crest (Fig. 1) should not be ignored [3‐5]. Even though the risk for suffering an iliac crest fracture after bone harvesting seems to be rather rare (0 to 4.1%) [3, 6‐9], it can alter the function for a few days up to a complete incapacity with the necessity for total bed rest or even surgical treatment [4]. Rare as well is the literature providing guidelines for the optimal location for iliac crest harvesting. Since 1995, just two experimental studies investigated this topic recommending that the graft should be harvested at least 30 mm posterior to the anterior superior iliac spine (ASIS) and the length of the graft should not exceed more than 30 mm to prevent fatigue iliac crest fracture [4, 10].

The objective of this investigation is to assess the risk for fatigue iliac crest fracture following cortico-cancellous bone harvesting at the anterior iliac wing.

The results of the FEA are explained by showing the stress distribution on the right iliac wing of five exemplary analysis. All stress values are in Pascal. Well-balanced stress values along the iliac wing and the absence of stress peaks are present for an intact iliac wing (Fig. 5a). Multiple stress peaks between the ASIS and the cutout can be detected when the gap in the iliac crest is placed 5 mm posterior to the ASIS (Fig. 5b–d). A lack of stress occurs along the iliac wing posterior to the gap. This area is of a triangular shape with one side being the edge of the cutout and one side being the iliac crest (Fig. 5b–d).

By successively placing the gap further away from the ASIS, the stress peak between the ASIS and the anterior edge of the cutout is notably reduced and the lack of stress along the iliac wing posterior to the cutout diminishes successively as well (Fig. 5c, d).

The stress distribution along the iliac wing is equal irrespective of the cutout size (15 × 20 mm versus 15 × 35 mm) but with a deeper cavity; stress values between the ASIS and the gap are notably higher. Contrary to the 15 × 20 mm cutout, a stress peak remains between the ASIS and the anterior edge of the 15 × 35 mm cutout when this is placed even 35 mm posterior to the ASIS. Furthermore, the area of the lack of stress posterior to the gap is significantly bigger for the 15 × 35 mm gap than for the 15 × 20 mm gap.

With a cutout that leaves the iliac crest intact, the stress distribution on the iliac wing between the cavity and the ASIS is notably relieved (Fig. 5e). There are no stress peaks on the iliac wing. The stress distribution posterior to the cutout is balanced contrary to every other model with a gap in the iliac crest.

Fractures of the iliac wing following autologous bone graft harvesting seem to be with less than 5% a rare complication [3, 6‐9] in comparison to the estimated overall morbidity rate of 19.37% [5]. In 2/3 of the cases, these fractures are fatigue fractures of the ASIS after harvesting the anterior iliac crest [12]. One of the main reasons for postoperative fractures of the ilium seems to be the sudden contraction of the muscles attached to the ASIS, sartorius muscle, and tensor fascia muscle [13]. To reestablish the function of these muscles as well as to induce pain relief to the patient in 16.6%, a surgical stabilization (Fig. 1) has to be performed [12].

Different surgical techniques including autologous rib transplantation [14] as well as xenogenous cancellous allografts [15] are proposed to restore the bone defect after iliac crest harvesting in order to prevent local complications. In our opinion, the method to cover the iliac gap by fixation of a plate [16, 17] (Fig. 6) is an adequate way to prevent fatigue fracture of the ASIS even though there is no clinical or biomechanical study to which factor the risk of an iliac crest fracture is reduced by it.

Looking at the result of our biomechanical FEA, the stress distribution along the iliac wing varies from critical to stress-relieved depending on location and size of the harvested bone. As expected, stress peaks occur especially between the ASIS and the anterior edge of the harvest site. The closer the harvest site is placed to the ASIS and the deeper the size of the harvest site is performed, the higher the stress peaks arise. The danger of a small crack spreading out from the anterior edge of the harvest site to the front of the ilium and thus causing a fracture of the anterior iliac crest becomes obvious.

Irrespective of the harvest size (15 × 20 mm versus 15 × 35 mm), the iliac wing does not show any significant stress peaks when the harvest site starts 20 to 25 mm posterior to the ASIS. Proving Hu et al. [4] right that bone graft taken 30 mm posterior to the ASIS significantly reduces the risk of anterior iliac crest fractures, this study even suggests going as close as 20 to 25 mm to the ASIS without a mentionable increase of the risk for anterior iliac crest fractures. Lengthening the harvest area from 15 to 30 mm does not notably impact the stress distribution between the ASIS and the cutout (data not shown). Nonetheless, the harvest site should be chosen as small as possible since many abdominal muscles are attached to the iliac crest and a herniation of visceral organs can be prevented.

Creating a gap in the iliac crest by bone harvesting interrupts the force flow that usually spreads out from the ASIS along the iliac crest to the posterior iliac wing. This way, a triangular shaped area of low stress posterior to the harvested site emerges (Fig. 5). Referring to Wolff’s law without the application of stress, this part of the iliac wing will become weaker [18], thus increasing the risk for a consecutive fracture at this area years after bone graft harvesting. Furthermore, it might be a reason for the most frequent complication caused by bone harvesting at the iliac crest: the persisting local pain [3‐6]. Whereas recent investigations show that iliac crest harvesting does not result in an increase of acute pain or narcotics consumption [19, 20], Sasso et al. point out that donor site pain remains a significant postoperative problem even years after treatment [21]. To avoid adverse events caused by autologous iliac crest, harvesting allografts as well as synthetic bone grafts always has to be considered as an alternative even though an evidence for superiority is not given yet [22, 23].

To eliminate the stress peaks between the ASIS and the harvest site as well as to preserve a force transmission from the ASIS to the posterior iliac wing, the bone graft could be harvested below the iliac crest (Fig. 5e). Having a continuous and uninterrupted iliac crest balances the stress distribution on the iliac wing with the implication that just a bicortical graft can be extracted. If a tricortical bone graft is needed, rebuilding the iliac crest with the fixation of a plate over the harvest area (Fig. 6) might recreate a balanced force distribution along the iliac wing. It should relieve the stress applied on the anterior iliac wing between the ASIS and the harvest site as well as eliminate the lack of stress in the area posterior to the harvest site.

Analyzing the influences on the iliac wing when maximum stress is applied by the muscles attached to the pelvis and by the forces that occur at the hip joint during gait analysis leaves possible outside influences out of consideration. Load applied on the pelvis by torsional mass is not taken into account. The points of attachment of muscles might vary from their real location since the forces are exported from an ideal model in AnyBody. Since this study focuses on the relative stress distribution on the iliac wing rather than the absolute stress values, the bone age of the used CT scan is not of significant relevance.

In conclusion, there are three options to significantly decrease the risk of fatigue iliac crest fracture following bone graft harvesting from the anterior iliac wing. First of all, leaving the iliac crest intact by harvesting below it significantly relieves stress peaks between the ASIS and the harvest site. It balances the stress distribution along the iliac wing as well. Secondly, a rather longer than deeper bone graft should be preferred from the biomechanical point of view. However, respecting the anatomy with many abdominal muscles attached to the iliac crest, a proper balance between length and depth of the bone graft needs to be found. Most of all a harvest site located at least 20 to 25 mm posterior to the ASIS should be preferred to minimize the risk of iliac fatigue fracture. Even though this study has not proven that a reconstruction of the iliac crest with plate fixation over the harvest area will prevent a fatigue iliac wing fracture, the authors are convinced of it.