Background
Giant cell tumors (GCTs) are benign but locally aggressive tumors with a relatively high rate of recurrence if not appropriately managed. GCTs involving the pelvis are extremely rare, accounting for only 1.5–6% of all GCTs of bones [
1,
2]. Treatment modalities include intralesional curettage with or without adjunctive procedures and wide resection [
2‐
5]. The options of curettage preserving the integrity of the pelvis can lead to an excellent functional outcome but a high recurrence rate, whereas wide resection has a low rate of recurrence but increases surgical morbidity with complications [
3,
4]. So, the management of pelvic GCTs involving the acetabulum remains a challenge for surgeons on how to balance the relative benefits of minimizing recurrence and maintaining postoperative hip function on account of their infrequency and the anatomic complexity of the pelvis [
6]. Owing to the local aggressiveness of GCTs, the initial surgical treatment is of vital importance for recurrence of the tumor in the pelvic region often makes it unresectable [
3]. So, even with the high rate of associated complications, wide resection is still recommended by several authors [
3,
6,
7].
The high rate of failure following traditional resection to region II of the pelvis is attributed to the extensive bone excised, no ideal implants achieving long-term stable fixation, and vulnerability to infection, etc. [
3,
6]. To decrease the risk of complications, some authors attempt to preserve the host bone as much as possible by minimizing the resection of healthy tissue surrounding the tumor. The technique “multiplanar osteotomy with limited margins” has been described by Avedian et al. [
8], who used this modality to successfully treat the selected patients with high-grade bone sarcomas by making angled bone cuts around a tumor; this preserves as much host bone as possible with the goal of minimizing bone and soft tissue ablation. Similarly, the technique of multiplanar osteotomy was successfully performed by Lam et al. [
9] and Gerbers and Jutte [
10] for periacetabular neoplasm in selected patients. However, there was no reconstruction for the resulting defect following resection of periacetabular bone; this may contribute to some hip instability, accelerated osteoarthritis, and postoperative hip dislocation [
9,
10]. Autogenous femoral head bone grafts have been widely applied for reconstruction of acetabular deficiency in cementless total hip arthroplasty (THA) for developmental dysplasia of the hip [
11,
12], and the long-term results are satisfactory. To our knowledge, there has been no such research regarding use of multiplanar osteotomy and reconstruction of autogenous femoral head bone grafts combined with cementless THA for pelvic GCTs involving partial acetabulum.
We performed the above technique in selected patients with pelvic GCTs involving partial acetabulum. The present study was to review the outcome of local recurrence, function, and any associated complications.
Discussion
In this study, a series of seven patients with pelvic GCTs involving ischium/pubis and partial acetabulum were carefully screened and treated by multiplanar osteotomy with a limited wide margin and reconstruction of autogenous femoral head bone grafts combined with cementless THA, with the goal of minimizing ablation of healthy bone and enhancing the hip function.
The treatment of pelvic GCTs involving acetabulum is usually difficult and controversial. Intralesional curettage can preserve the integrity of the pelvis and obtain good hip function. However, a systematic review reported that those who had undergone intralesional surgery had a higher rate of local recurrence (33.3%) than those treated by wide resection (2%) [
6]. Traditional options of wide resection are to remove the whole bone of region II, which destroy the integrity of the pelvis. A variety of complications such as wound infection, nonunion, bone absorption, deep infection, or hardware failure are obvious; this leads to poor hip function. Until now, few publications have specifically addressed pelvic GCTs involving the region III with partial acetabulum extension [
2,
4,
5,
7,
16‐
20] (Table
2). It remains a challenge to musculoskeletal oncologists on how to balance minimizing the recurrence rates and maximizing the hip functional outcomes.
Table 2
Demography of patients with pelvic GCTs involving region III with partial acetabular extension in various studies
| IA (1) | ILC + phenol + allograft | Migration | None | None |
| IA (3); PA (1) | ILC + autograft | Infection (2) | 3 | 2 |
| IA (1) | ILC + cementation | None | None | None |
| IPA (1); IA (1) | ILC + cryosurgery + prosthesis (1); ILC + cryosurgery + radiation + cementation (1) | Sciatic nerve palsy (1) | None | None |
| IA (2); PA (2); IPA (2) | ILC + cementation (3); ILC + cryosurgery + bone graft + radiation (1); ILC + hip transposition (1); sole radiation (1) | Screw dislocation (2); femoral head necrosis (1); subluxation of femoral head (1) | None | None |
| IA (3) | WR + iliofemoral arthrodesis | None | None | None |
| PA (1) | WR + iliofemoral arthrodesis | None | None | None |
| IPA (1); IA (1) | WR + iliofemoral arthrodesis (1); WR + THA (1) | Opening of ilum (1); infection (1) | None | None |
Mnaymneh and Mnaymneh [ 20] | IPA (1) | Wide amputation | None | None | None |
Current study | IA (6); PA (1) | WR + autograft + THA | None | 1 | None |
In order to reduce the local recurrence rate and improve the hip function, wide resection should be performed; meanwhile, traditional wide resection of region II should be avoided. In a study of five patients with malignant neoplasms of the periacetabular region, Lam et al. [
9] performed acetabular-preserving resections that preserved the weight-bearing acetabulum, and satisfactory outcomes were obtained with a median follow-up of 37 months. Gerbers and Jutte [
10] reported one patient with chondrosarcoma of the region III treated by partly resecting the frontal part of the acetabulum with computer assistance to obtain a safe margin and achieved an excellent postoperative function with a follow-up of 3.5 years. In the present study, cases were carefully screened to determine whether the patients were candidate for this type of surgery. With the assistance of computer simulation, we make precise preoperative planning carefully. The planes of osteotomy around the acetabulum were created to make sure a safe margin was achieved when excising the partial acetabulum. At the same time, enough host bone can be preserved for reconstruction. It is of particular importance that the dome of the acetabulum should be retained after osteotomy. Generally, tumors extending proximally beyond the supra-acetabular line are excluded in our study. This method can avoid complex reconstruction, and leads to better hip function than the standard resection of region II. In our series, limited and safe margins are achieved by multiplanar osteotomy with the help of precise preoperative simulation. How to achieve a limited and safe margin should be considered. Computer-assisted surgery, a trustworthy means of navigation, has been reported in resections of pelvic tumors [
10]. This can provide precise imaging and achieve desired safe margins. Lam et al. [
9] reported that the precise planning of the resection was carried out with computer navigation software. In the current study, we use the Mimics software to visualize and segment the CT images and render 3D pelvic bone. The tumor 3D models can be extracted. Then, the relationship between the tumor and the healthy bone can be obviously displayed. Accurate planes of osteotomy are created in the window of the 3D view. Taking the apex of acetabulum and acetabular fossa as reference during operation, the senior surgeon can easily achieve limited and safe margins with the guide of preoperative simulation. Postoperative biopsy further confirmed a clear margin.
During preoperative planning, it is of vital importance to consider the residual defect following resection and the material that will be utilized to fill this defect. Bulk autogenous graft from the femoral head has been widely used to fill the deficient acetabula in patients with developmental dysplasia of the hip. In patients with the use of cemented acetabular components, the failure rate ranges from 38 to 46% over at least a 10-year follow-up [
21,
22], mostly because of asymptomatic loosening or graft collapse. With the advent of cementless acetabular components, the 10-year survival rate without acetabular revision for any reason ranges from 94 to 100% [
11,
12,
23]. Certainly, one must be aware that the size of the defect that is filled by the graft may affect survivorship. Several authors recommended that the coverage of the socket by the graft not exceed ranging from 30 to 50% [
11,
12,
24]. In this series, only the patients with > 50% preserved host bone following resection of the periacetabular region are included. At the last follow-up, the grafts are union and no acetabular components needed to be revised. The character of our patients differs from that of other reports. So, we cannot make any comparison. Our clinical experiences indicate that as long as one of the two columns is retained and the resulting defect does not exceed the supra-acetabular line, this is a viable method of reconstruction for patients with pelvic GCTs involving partial acetabulum.
Several factors can influence the successful incorporation of the autogenous femoral head bone grafts: First, graft orientation in relation to the host bone is of significance. We always make the portion of the femoral neck in contact with the proximal deficient acetabulum, and the femoral head faces the acetabular fossa. The cortex of the femoral neck can bear greater pressure and provide better holding force when implanting the screws, while the cancellous bone of the femoral head can be easily trimmed and reamed when preparing the socket. Second, it may be technically demanding to match the defect. According to the preoperative simulation, we had a preliminary understanding on how to trim the bone graft. Then, the reciprocating bone saw was well applied to carefully trim the graft to obtain the satisfactory matching of the defect. Third, it is of particular importance that the screws cannot protrude into the socket following reaming of the acetabular fossa. And screw orientation should be close or parallel to the conduction force of the acetabulum for the reason that axial compression of the graft can enhance bone graft incorporation with the host bone. Our preoperative simulation to the placement of the screws can be a good solution to this concern.
In this series, six out of seven patients were grade III lesions at presentation. The probable reasons of the high rate are uncharacteristic clinical presentation, easily confused with low back pain, arthritis, muscle strain, or lumbar intervertebral disc herniation, no visible swelling, and misinterpreted radiographs caused by gas-filled intestines. In the present study, local control was achieved in six out of seven patients. Recurrence occurred in one patient, confirmed by further MRI and open biopsy. It should be noted that this is a grade III lesion, and the recurrent tumor located not in the periacetabulum but in the soft tissue. We believe it might not be possible that no safe margins in the periacetabular region were achieved. The possible reason of recurrence may be inadequate soft tissue margin or the presence of satellite lesions that were unable to be seen by the naked eye.
Several limitations of our study should be noted. First, only a small proportion of cases with pelvic GCTs involving ischium/pubis and partial acetabulum were candidates for this type of surgery. It may have led to deviation over the results of local recurrence, bong graft healing, or failure of acetabular components. Second, although we made a precise preoperative planning relying on computer software to determine the plane of osteotomy, it might include a somewhat subjective process such as the surgeon’s experience with osteotomy. It may be more accurate with the assistance of the osteotomy guide in the future. Third, the follow-up is short; additional local recurrences may be detected with longer follow-up. Nevertheless, 70% of local recurrences occur within 2 years [
25]. Fourth, in this retrospective series, no patient was treated with denosumab. This drug is a human monoclonal antibody against the receptor activator of nuclear factor kappa-B ligand, which can shrink the size of large GCTs to facilitate tumor resection [
26]. So, it should be strongly recommended that denosumab be subcutaneously administered before surgery, especially in patients with grade III.