First application of three-dimensional designing total hip arthroplasty with long uncemented stem for fibrous dysplasia patients combined with hip joint osteoarthritis

Fibrous dysplasia (FD) is a benign lesion characterized by the replacement of normal bone tissue with immature fibro-osseous tissue leading to widening and thinning of the cortex. It can occur in single bone (monostotic) or multiple bones (polyostotic) [1]. The proximal femur, which is the most common site, can gradually form progressive coxa vara and bowing deformity under the continuous mechanical stress and repeated fractures, and result in functional impairment and pain [2–4].

In a large proportion of these patients, secondary hip joint osteoarthritis (OA) can be caused by the abnormal mechanical alignment of the proximal femur [5, 6]. Besides, the severity of the deformity and the presence of FD lesions in the peri-articular bones (head of femur and acetabulum) were proven to be independently and closely associated with the development of secondary hip joint OA in these patients [7]. Although, the reported incidence of secondary hip joint OA in patients with proximal femur FD was 13% [7], with the disease progression, the final incidence will highly increase.

The surgical treatment for proximal femur FD combined with hip joint OA remains challenging due to the poor bone conditions, severe bone deformity, dysfunctional hip abductor muscle and non-standard implant choice. Recently, some literatures have focused on this problem, total hip arthroplasty (THA) has been accepted as an effective method [6]. The short uncemented stem or cement fixation were applied, however, the postoperative prosthesis loosing was severe, long-time fixation of femoral stem was unsatisfactory, and the number of cases was limited [6, 8, 9].

3D (three-dimensional) image technology has opened up a new era to this problem. In order to preserve the remaining femoral bone stock and attain a more precise, individual and durable prosthesis implantation, the 3D designing THA with long uncemented stem was firstly applied in our institution. This study attempts to present the detailed design procedures, perioperative management and evaluate short-term clinical outcomes of this novel therapeutic method.

In our series, seven patients received 3D designing combined osteotomy and THA, and the other five patients underwent 3D designing THA. The mean surgical time was 150 (range, 125–190) min. The average amount of intraoperative bleeding was 730 (range, 550–1350) ml. There was no infection, dislocation, thromboembolism, postoperative wound problems or mechanical failures.

The average follow-up was 47 (range, 35–56) months (Table 1). All patients had normal functions of the knee joint, ankle joint and foot, and satisfactory function of hip joint. At the last follow-up, 7 cases were painless, 5 patients had significant relief of pain. All cases were able to walk, even though 3 cases walked with the aid of cane during a long walk. Only one patient got residual deformity of lower limbs because of the severe bone condition preoperatively, but the function outcome for this patient was satisfactory.

The average Harris Hip Score was improved from 46.08 (range, 13–67) points preoperatively to 93.72 (range, 83–100) points at the last follow-up (Table 1). The average modified criteria of Guille was improved from 3.2 (range, 1–7) points preoperatively to 8.6 (range, 6–10) points at the last follow-up. Results were excellent in 7 patients (Figs. 5 and 6), good in 4 patients, fair in 1 patient, and poor in 0 patients. The details of clinical outcomes were shown as Table 2.

Categories	Unsatisfactory		Average		Satisfactory
	Preoperative	Postoperative	Preoperative	Postoperative	Preoperative	Postoperative
Pain	6	0	5	5	1	7
Hip motion	7	0	3	3	2	9
Limping	7	0	4	4	1	8
Activities of daily living	6	0	4	3	2	9
Social activities	5	0	4	2	3	10

Notes: 12 femora were evaluated by five different categories. Hip function: unsatisfactory scored 0 points, average scored 1 point, satisfactory scored 2 points. In a case with a maximum score of 10 points, ≥9 points was identified as excellent, 7 or 8 points as good, 5 or 6 points as fair, and < 5 points as poor

All of the patients were divided into two groups according to whether or not received osteotomy, including seven patients received 3D designing combined osteotomy and THA and five patients underwent 3D designing THA only. For the seven patients received 3D designing valgus osteotomy, the limb-length discrepancy was corrected from 3.8 (range, 1.5–7) cm preoperatively to 1.0 (range, 0–3) cm postoperatively. The mean extremity lengthening was 2.8 (range, 1.5–4) cm (Figs. 1, 4 and 7). No postoperative neurovascular complications or dysfunction of muscle was found owing to the lengthening of the lower limb. The bone union was achieved in all patients, and the average healing time was 4.2 (range, 3–6) months. The details of Harris Hip Score were shown as Table 3. Significant differences were only detected in limp (p = 0.020) between the two groups. But there were no significant differences in pain (p = 0.930), distance walked (p = 0.763), support (p = 0.763), stair climbing (p = 0.763), sitting (p = 1.000), shoes and socks (p = 1.000), transportation (p = 1.000), deformity (p = 0.424), range of motion (p = 0.763), and Harris Hip Scores (p = 0.402) between the two groups (Table 3).

Case	Pain	Distance walked	Support	Limp	Stair climbing	Sitting	Shoes and socks	Transportation	Deformity	Range of motion	Harris scores
1	40	8	7	8	2	5	4	1	4	4	83
2	44	11	11	8	4	5	4	1	4	5	97
3	40	11	11	11	4	5	4	1	4	5	96
4	44	11	11	11	4	5	4	1	4	5	100
5	44	11	11	8	4	5	4	1	4	5	97
6	40	11	11	8	4	5	4	1	4	5	93
7	40	8	7	8	2	5	4	1	4	4	83
8	40	11	11	11	4	5	4	1	4	5	96
9	44	8	7	8	2	5	4	1	0	4	83
10	44	11	11	8	4	5	4	1	4	5	97
11	44	11	11	11	4	5	4	1	4	5	100
12	44	11	11	11	4	5	4	1	4	5	100

Notes: Harris Scores range: 0–100

For FD combined with hip joint OA patients, especially who have severe proximal femur deformity, corrective osteotomy and THA are both needed. Corrective osteotomy can solve local deformity and re-establish mechanical alignment of the lower extremity [1]. THA can relieve pain and improve the limb function [6]. Some previous publications have demonstrated the value of application of computer-assisted technology, robotics-assisted technology and rapid prototype model in complex hip surgery. One review has summarized computer-assisted technology, robotics-assisted technology can help surgeons perform accurately to achieve the clinical objectives in complicated THA [13]. In 2013, Won et al. reported that the pelvic rapid prototype model was used in 21 complex hips to simulate the implantation of the acetabular prosthesis preoperatively, the implantation of acetabular component was successful and 80.9% of the used acetabular component was within 2 mm of the predicted size [14]. In 2016, Inaba used 3D preoperative planning and intraoperative navigation for rotational acetabular osteotomy in acetabular dysplasia patients, the rotational acetabular osteotomy could be performed more accurate and safer [15]. But the 3D preoperative design is still rare applied in THA with severe proximal femoral deformity. In order to improve the perioperative accuracy rate of corrective osteotomy and THA, 3D preoperative design was firstly used in FD combined with hip joint OA patients. This study demonstrated the details of 3D preoperative design and perioperative management, and the short-term outcomes of THA with long uncemented stem for this kind of patients.

With the progression of proximal femur FD, coxa vara deformity is inevitable. For this deformity, valgus osteotomy can improve the limb function and prevent fractures via re-establishing neck-shaft angle and reconstructing the lower limb strength line [16]. The ways of valgus osteotomy are based on the deformity classification. Ippolito proposed a classification method including six types in 2014 [17]. However, Ippolito’s classification has a major limitation, which is the surgeons cannot determine osteotomy site or internal fixation method according to this classification. In 2015, our institution reported a novel method of radiographic classification including five types according to three parameters: neck-shaft angle, varus deformity of the proximal femoral shaft, and reduction in proximal femoral strength [10]. In our series, according to our institution’s classification, there were two type 2 patients, three type 3 patients, three type 4 patients, and four type 5 patients. Femoral valgus osteotomy was done in the three type 4 and four type 5 patients, who had varus deformity in the proximal femur shaft. Nevertheless, the current classifications, including ours, only evaluate the femoral deformity but not the injury degree of the ipsilateral hip joint [10, 17]. Moreover, valgus osteotomy can lengthen the lower extremity to some extent. Previous researches demonstrated that corrective osteotomy can lengthen extremity by 2.3–3.4 cm on average [16, 18]. In our study, the lower limb has been lengthened by an average of 2.8 (range, 1.5–4) cm.

Current classifications of proximal femur FD are only based on the images of X-ray and CT, but not a multi-dimensional evaluation [10, 17]. To our best knowledge, there is no report using 3D designing femoral reconstruction to guide the operation. Additionally, the proximal femur medullary cavity is varied in diameter and the bone condition is poor for FD patients, the implantation of prosthesis stem is challenging. But it is still very rare to pay attention to the bone condition of the affected proximal femur, especially the surface of medullary cavity. The 3D reconstructive images can fully and precisely evaluate the patients’ femoral deformity sites, affected bone length, abnormal mechanical alignment, bone condition, and multi-level diameter of the medullary cavity. Our results revealed that 3D designing osteotomy did not reduce the hip joint function when comparing the patients only receiving THA and the patients receiving osteotomy and THA. The preoperative 3D design can not only make better plan of the osteotomy but also simulate the feasibility of implanting long prosthesis stem more accurately. During this procedure, we can confirm the length and diameter of the prosthesis stem before surgery, so that the surgical difficulties can be sharply decreased, and the mean operating time and the average amount of intraoperative bleeding can be controlled well.

Primary stability and long-term survival are the most important factors to consider when choose the prosthesis stem, especially for these patients with bone deformity and poor bone condition. Firstly, uncemented stem was chosen in our series. The reasons has been discussed that the affected bone characterized with the wide and irregular medullary cavity and thin cortex, is not strong enough to support the cement-bone interface [19, 20]. Additionally, due to the relatively massive intramedullary bleeding, uncontrollable bleeding will undoubtedly influence the cement-bone interface during cemented stem implantation [6]. What’s more, for patients who require femoral osteotomy, the cement fixation may affect osteotomy healing. Secondly, long stem should be applied. Sierra reported that proximal femur FD patients who received short uncemented stem fixation showed a high rate (60%) of early aseptic loosening, and long stem prosthesis was highly recommended for proximal femur FD patients combined with hip joint OA. Besides, they suggested that the stems length should exceed the lesion length by at least 2 femoral canal diameters to decrease the risk of postoperative fractures [6]. Some data demonstrated that besides long stem, increasing bone mass by cortical bone or compressive bone grafting and custom-made megaprosthesis can also improve the integral strength of the proximal femur [21]. The same theory has been accepted during femoral revision surgery after ordinary THA [22, 23]. What’s more, in some rarely extreme cases, if the usual size of the femoral stem (solution revision femoral prosthesis) was not appropriate for the patient, the customized femoral stem could be produced assisted by 3D design preoperatively.

The surgical skill of the stem implantation is another crucial point worthy of notice. First, precision osteotomy assisted by 3D designing customized osteotomy guides should be performed and micropendulum saw was required to be used to reduce bone loss and cortical injury. Second, The lesion required to be removed thoroughly before reaming, which would prevent poor osteointegration after uncemented fixation [6]. Third, the femoral cavity needed to be reamed before the implantation of long stem. 3D designing preoperative plan and intraoperative X-ray fluoroscopy played a very important role in the precise reaming of the bone cavity and preventing cortical bone damage. Fourth, in order to increase initial stability, the diameter of reaming needed to be slight (0.5 mm) smaller than the diameter of the stem because of the flexibility of the cortical bone with FD. Fifth, during stem implantation, the anteversion angle of the stem should be pay attention to, because the anteversion angle of the host proximal femur is always less than normal. Sixth, the procedure of wire binding was usually used to re-establish abductor muscle and prevent intraoperative periprosthetic fractures [24]. During osteotomy and prosthesis implantation, periosteal should be protected to promote bone healing. Finally, massive impaction of cancellous allograft should be used before the implantation of acetabular cup and prosthesis stem. Massive impaction allograft is the critical procedure to improve the potential of osteointegration and to prevent uncemented implant loosening [16, 25, 26].

It should be noted that this is a retrospective and short-term follow-up study related to 3D designing THA with long uncemented stem for FD patients combined with hip joint OA. Additionally, this study included a limited number of patients. Nevertheless, this study still demonstrates the clinical value and potential of 3D designing osteotomy and implantation of long uncemented stem for FD patients combined with hip joint OA.

The 3D designing THA with long uncemented stem seems to be a reliable method for FD patients combined with hip joint OA. To our best knowledge, this study includes the largest number of patients using long uncemented stem and is the first application of 3D design technology for osteotomy and stem implantation currently. The 3D designing corrective osteotomy can re-build the normal physiological structure of the affected bone accurately and improve the lower limb function. The 3D designing implantation of long prosthesis stem is more precise, contributes to primary stability, and provides necessary condition for long-term survival. Moreover, perioperative management should be carefully paid attention to. Nevertheless, the long-term outcomes and larger cases results are still required.