Fixation stability and implant-associated complications in periacetabular osteotomy: a comparison of screw and K-wire fixation

Periacetabular osteotomy (PAO) is an established technique in the treatment of developmental dysplasia of the hip (DDH) in young adults [1‐3].With the aim of preventing or at least delaying development of secondary osteoarthritis, excellent results both clinically and radiologically have been described in short-term, mid-term and long-term follow-up [1‐4].

The surgical technique has been well described [1]. During PAO, the acetabulum is completely released from the pelvis through five osteotomies resulting in free movement and the possibility of three-dimensional reorientation of the hip socket [5]. In the original description of the procedure, fixation of the reoriented acetabular fragment is achieved through screw fixation [1, 5]. According to the current literature, screw fixation is the standard procedure in fragment fixation in PAO. But despite good clinical results, non-union remains a common complication [6‐8]. The literature shows that with increasing rigidity of the fragment fixation, the rates of non-union also increase [9‐11]. In a study by Clohisy et al., screw fixation was performed and 8% non-unions were observed [10]. In another study by Clohisy et al., screw and plate fixation was combined and the rate of non-union increased to 19% [9]. As an alternative, semi-rigid Kirschner wire (K-wire) fixation of the acetabular fragment combined with allogenous bone grafting has been described showing good results with a reduced rate of non-union [11]. The idea behind this technique is to combine a semi-rigid fixation with a stable press-fit allograft allowing minimal movement of the acetabulum, but enough stability to prevent a loss of correction. With less fixation rigidity, one could assume that a loss of correction is more likely. However, comprehensive studies comparing K-wire fixation and screw fixation regarding stability and complications are lacking.

This study aims to assess the fixation stability and rate of implant-associated complications of K-wire fixation compared to screw fixation after acetabular reorientation in PAO.

DDH is a complex pathology of the hip and the leading cause of secondary osteoarthritis of the hip [15]. If diagnosed in time and correctly indicated, PAO is a good option in operative therapy of DDH in the adult showing good to excellent outcomes [2, 4, 16‐18].

Through three-dimensional reorientation and fixation of the acetabulum, better coverage of the femoral head is provided. So far, most previous studies described the use of screw fixation. It is assumed that this rigid fixation is necessary to provide and maintain stability and thus secure consolidation of the osteotomies [1, 4, 18, 19]. The idea behind fragment fixation with K-wires is that combining a semi-rigid fixation with a stable press-fit allograft allows minimal movement of the acetabulum, yet still provides enough stability to prevent a loss of correction.

To our knowledge, this is the first study to compare the stability of K-wire- and screw fixation in PAO.

The most important finding of the presented study is that there is no clinically significant loss of correction with the investigated modified technique using K-wire fixation compared to screw fixation. It can therefore be assumed that even if semi-rigid fixation is used, sufficient stability is achieved to obtain the correction result. Even a more restrictive postoperative treatment in the sense of prolonged partial weight-bearing did not seem necessary according to the results of our cohort, since all patients were mobilized according to the same postoperative mobilization regimen.

A previous study investigated the occurrence of non-unions in PAO with K-wire fixation. However, detailed radiological evaluation of loss of correction was not reported in that study [6]. Other studies comparing different screw-fixation techniques and their stability have been published numerously. These studies involved the biomechanical comparison of different fixation models in PAO. [20‐23]. For example, Babis et al. were able to show higher stability in more rigid fixation constructs in an in vitro model [23]. Widmer et al. were also able to show higher stability for more rigid fixations [21]. However, these studies were in vitro models with simulated forces similar to or even exceeding those of full weight-bearing. Furthermore, none of these studies investigated the stability of K-wire fixation in PAO.

K-wire fixation as an alternative to screw fixation has previously been reported in pelvic surgery. It is successfully used for fixation of pelvic osteotomies in children and adolescents and also, for example, in pelvic osteotomies for Perthes disease [24]. However, the conditions in these cases can only be applied to the biomechanical conditions in PAO in adults to a limited extent. Biomechanical studies comparing fixation with K-wires and screws in complex multidimensional pelvic osteotomies are lacking.

Our study involved in vivo conditions using a mobilization regimen with postoperative partial weight-bearing. Postoperative mobilization with partial weight-bearing is regularly used after PAO and is also described in the literature by the first describers of PAO [1, 5].

In terms of complications, the K-wire group showed significantly more implant-associated soft tissue irritations requiring implant removal compared to the screw-fixation group. In our opinion, this is caused by the relatively bigger portion of osteosynthesis material protruding from the bone of the iliac crest in an area with relatively little soft tissue coverage. A previous study found similar complication potential in their study without further quantifying or comparing it [11]. Thawrani et al. described a similar incidence of soft tissue irritation in their study of complications of the established technique with screw fixation [25], similar to the screw group investigated in this study. On the one hand, retained implants may pose a problem for MRIs that may be necessary in the future, as metal implants cause artifacts and thus reduce the quality of the imaging [26].

On the other hand, retained implants may pose a challenge for conversion to THA, which is still necessary in a relevant proportion of patients [4].

There was no significant difference in implant failure between the two groups. No major complications such as non-union or intraarticular implant migration were found in either group. Even though in our study little minor and no major complications can be reported, PAO is still a surgical method in which higher complication rates were frequently described. Overall complication rates in previous studies vary between 1 and 86% for major complications and 11 and 100% for minor complications [11, 25].

The K-wire group had significantly reduced surgery time compared to the screw group. The fact that the two surgeons each used one of the fixation techniques results in a certain bias and the reduced operation time cannot be attributed with certainty to the choice of fixation. However, from our point of view it seems logical that the operation time can be reduced if the intermediate step of changing the temporary fixation with K-wires to the definitive fixation with screws is omitted if no screws are used, whereas in screw fixation this intermediate step is regularly performed [1]. This could be an advantage insofar as various studies have shown that prolonged duration of surgery is a risk factor for perioperative complications in orthopedic surgery as well as in other surgical fields [27, 28]. Tahwrani et al. found that prolonged surgery time in PAO is a risk factor for major complications as well as blood loss in PAO and therefore needs to be avoided as much as possible [25].

This study has several limitations. First, it has a retrospective study design. Second, preoperative measurements of the relevant radiological parameters show more dysplastic values in the K-wire group. However, both groups were corrected to almost the same extent and the desired correction was achieved as shown in the postoperative measurements in both groups. Third, in our cohort the procedure was performed by two different surgeons. Fourth, 3 months follow-up is relatively short. However, we believe that with painless full weight-bearing and radiologically proven bony consolidation already achieved at this time, further loss of correction thereafter seems unlikely.

This study demonstrated that K-wire fixation is a viable and safe option in fragment fixation in PAO with similar stability and complication rates compared to screw fixation. An advantage of the method is the significantly reduced operating time. A disadvantage is the significantly higher rate of implant-associated soft tissue irritation requiring implant extraction. Further studies are needed to examine the presented surgical technique at a long-term follow-up. Furthermore, biomechanical studies could provide further information about the stability of the investigated method. These findings can be taken into account by surgeons when choosing the form of fixation of the acetabular fragment in PAO.