The functional and oncologic results of eighteen patients with primary malignant periacetabular tumors were reviewed to determine the impact of surgical treatment. The reconstruction procedures were endoprosthesis (11), hip transposition (4), iliofemoral arthrodesis (2), and frozen bone autograft (1). After a mean follow-up of 62 months, 13 patients were alive and 5 had died of their disease; the 5-year overall survival rate was 67.2%. The corresponding mean MSTS scores of patients with endoprosthesis (11) and other reconstructions (7) were 42% and 55% (49%, 68%, and 50%), respectively. Overall, postoperative complications including deep infection or dislocation markedly worsened the functional outcome. Iliofemoral arthrodesis provided better function than the other procedures, whereas endoprosthetic reconstruction demonstrated poor functional outcome except for patients who were reconstructed with the adequate soft tissue coverage. Avoiding postoperative complications is highly important for achieving better function, suggesting that surgical procedures with adequate soft tissue coverage or without the massive use of nonbiological materials are preferable. Appropriate selection of the reconstructive procedures for individual patients, considering the amount of remaining bone and soft tissues, would lead to better clinical outcomes.
Although primary sarcomas arising in the pelvis are relatively rare, surgical treatment for these diseases remains difficult. Resections of pelvic bone are classified according to the system of Enneking and Dunham (Figure
Diagram showing the resected area according to the classification system of Enneking and Dunham.
Over the past 10 years, a variety of reconstructive procedures have been employed at our institution. The purpose of the present study was to evaluate the surgical outcome, including complication rate and functional score, in patients after acetabular resection according to the surgical procedures employed, and to determine the clinical and functional outcome after resection of periacetabular tumors.
We retrospectively reviewed 18 patients with primary periacetabular bone tumors who underwent acetabular resection and reconstruction between June 1996 and December 2012. Their clinical data, treatment modalities, and treatment outcome were reviewed retrospectively by reference to the medical records. The following data were examined: demographic data (patient age at operation, gender, tumor size, and histologic diagnosis), surgical details (reconstructive procedures, lesion resected, and surgical margins), lesion resected, adjuvant therapy (chemotherapy and radiotherapy), postoperative complications (e.g., infection or dislocation), oncologic outcomes, and functional outcomes.
Acetabular lesions were resected using a variety of procedures according to the classification system of Enneking and Dunham (Figure
Function was assessed at the final follow-up using the MSTS system developed by Enneking et al. [
Ethical approval was obtained from the Institutional Review Boards of National Cancer Center Hospital.
Patient demographics and treatment data are summarized in Table
Patient characteristics for the entire study population and surgical/oncological outcome.
Description | Number |
---|---|
Patients | |
Male 13, female 5 | Total 18 |
Age (at diagnosis) | 41 years (8–69) |
Tumor size | 11.7 cm (7–20) |
Diagnosis | |
Osteosarcoma | 8 |
Chondrosarcoma | 5 |
MFH of bone | 2 |
Ewing sarcoma | 2 |
Fibrosarcoma of bone | 1 |
Neoadjuvant therapy | |
Polychemotherapy | 11 |
Radiotherapy | 0 |
Adjuvant therapy | |
Polychemotherapy | 9 |
Radiotherapy | 1 |
Follow-up | 62 months (8–155) |
Resected area (Enneking classification) | |
PII | 2 |
PI-II | 4 |
PII-III | 8 |
PI-II-III | 4 |
Surgical outcome | |
Wide margin | 17 |
Marginal margin | 0 |
Intralesional margin | 1 |
Oncological outcome | |
No evidence of disease (NED) | 11 |
Alive with disease (AWD) | 2 |
Dead of disease (DOD) | 5 |
Prognosis | |
Overall survival (five years) | 67.2% |
The acetabular lesions were resected using a variety of procedures according to the classification system of Enneking and Dunham [
The surgical margins in this study group were classified as wide in 17 patients and intralesional in 1 patient (Table
After a mean overall follow-up of 62 months (range, 8–155 months), 11 patients (61%) had no evidence of disease (NED), 2 (11%) were alive with disease (AWD), and 5 (28%) had died of the disease (DOD) (Table
Cumulative overall survival curve for all patients estimated by the Kaplan-Meier method.
Postoperative complications are listed in Table
Complications according to surgical treatment.
Surgical procedures | Number of patients | Number of patients with major complications | Complications (number) | Local recurrence (number) |
---|---|---|---|---|
Endoprosthesis | 11 | 6 (55%) | Deep infection ( |
1 |
Other reconstructions | 7 | 1 (14%) | 0 | |
Hip transposition | 4 | 1 (25%) | Deep infection ( |
0 |
Iliofemoral arthrodesis | 2 | 0 | Implant breakage ( |
0 |
Frozen autograft | 1 | 0 | Osteoarthritis ( |
0 |
|
||||
Total | 18 | 8 (44%) |
|
1 |
After endoprosthesis replacement, most patients suffered postoperative complications. Among 11 patients who underwent endoprosthesis reconstruction, 6 (55%) had major complications, which required surgical interventions. The complications comprised superficial infection in 2 patients (18%), deep infection in 2 (18%), wound complication in 3 (27%), dislocation in 2 (18%), abdominal hernia in 1, and local recurrence in 1. Eight additional surgical procedures were performed in patients of this group, including 6 revisions for deep infection or wound complication and 2 for implant dislocation. No complications developed in 5 patients whose gluteus medius or gluteus maximus was preserved, or in those who underwent coverage of the large soft tissue defect with a rectus abdominis myocutaneous flap [
On the other hand, among those who performed other reconstructions than endoprosthesis, only one patient (14%) experienced major complication, which required surgical intervention. After hip transposition in 4 patients, one suffered postoperative infection, and all of the patients had leg-length discrepancy. One patient required additional surgical procedure for deep infection and wound problem. This patient underwent reconstruction with a Gore-Tex sheet around the iliac resection site, which was considered to have caused postoperative infection.
Of the 2 patients who underwent iliofemoral arthrodesis, both had leg-length discrepancy and one suffered implant complication. One patient suffered screw breakage at 2 years after surgery, which required an additional fixation. Later the patient underwent a limb-lengthening operation for leg-length discrepancy at another institution. The other patient died of metastatic disease at 2 years after surgery.
One patient who underwent reconstruction with a frozen bone autograft showed delayed postoperative wound healing. At 1 year after surgery, the patient presented with hip osteoarthritis, which has since shown gradual progression.
The mean functional score in the present series, according to the MSTS system, was 14.2 points (47%) at the latest follow-up. The functional results according to the various surgical procedures employed are shown in Table
Functional outcomes according to surgical treatment.
Surgical procedure | Score according to musculoskeletal tumor society (MSTS score) system | ||||||
---|---|---|---|---|---|---|---|
Pain | Function | Acceptance | Support | Distance | Gait | Total | |
Endoprosthesis | 64 (3.2) | 42 (2.1) | 46 (2.3) | 6 (0.3) | 44 (2.2) | 42 (2.1) | 42 (12.5) |
Other reconstructions | 82 (4.1) | 54 (2.7) | 57 (2.8) | 17 (0.8) | 57 (2.8) | 60 (3) | 55 (16.4) |
Hip transposition | 90 (4.5) | 50 (2.5) | 70 (3.5) | 0 (0) | 35 (1.75) | 50 (2.5) | 49 (14.7) |
Iliofemoral arthrodesis | 80 (4) | 60 (3) | 50 (2.5) | 60 (3) | 90 (4.5) | 70 (3.5) | 68 (20.5) |
Frozen autograft | 60 (3) | 60 (3) | 20 (1) | 0 (0) | 80 (4) | 80 (4) | 50 (15) |
|
|||||||
Mean scores | 73 (3.6) | 48 (2.4) | 51 (2.5) | 10 (0.5) | 49 (2.5) | 49 (2.5) | 47 (14.2) |
The 11 patients who underwent endoprosthetic reconstruction had a mean functional score of 12.5 points (42%) out of a maximum of 30 points (range, 13–70%). All patients were unable to walk without walking aids; no patient had a score of >3 points (out of a possible 5 points). Eight patients required constant pain medication with nonnarcotic analgesics (four with moderate pain and four with mild pain). The mean emotional acceptance score was 46% (range, 0–60%). Notably, the mean MSTS score for the six patients without postoperative deep infection or dislocation was 55% (Figure
(a) Preoperative anteroposterior radiograph of the pelvis of a 28-year-old female who had osteosarcoma of the acetabulum and ilium. ((b) and (c)) Gadolinium-enhanced axial T1-weighted MRI showing the tumor arising in the acetabulum and ilium. After neoadjuvant chemotherapy, the patient underwent tumor wide resection and endoprosthetic reconstruction with no postoperative major complications. (d) Postoperative radiograph 3 years after endoprosthetic reconstruction.
The 4 patients who underwent reconstruction with hip transposition had a mean functional score of 49% (range, 33–67%). They complained of the least amount of pain. Only one patient required continuous use of analgesic medications. Notably, the mean emotional acceptance score was 70% (range, 40–100%), which was the highest among all the groups. All of these patients had limb-length discrepancy (Figure
(a) Preoperative anteroposterior radiograph of the pelvis of an 8-year-old boy who had Ewing sarcoma in the ilium and acetabulum. (b) Axial T2-weighted MR image showing the tumor arising in the acetabulum. (c) Postoperative radiograph one year after hip transposition.
Two patients who underwent iliofemoral arthrodesis had a mean functional score of 68%. One patient had no pain, and the other had mild pain. The mean emotional acceptance score was 50%. One patient required no support for walking (Figure
(a) Preoperative anteroposterior radiograph of the pelvis of a 14-year-old boy who had Ewing sarcoma of the acetabulum and pubis. (b) Axial T2-weighted MRI showing the tumor arising in the acetabulum. (c) Postoperative radiograph showing plate fixation of the proximal femur to the remaining ilium after PII-III resection. (d) Plain radiograph showing screw breakage 2 years after first iliofemoral arthrodesis. (e) Plain radiograph 11 years after refixation.
The patient who was reconstructed with a frozen bone autograft had a functional score of 50%. He complained of moderate pain, which seemed to be caused by osteoarthritis, and used nonnarcotic analgesics (Figure
(a) Preoperative anteroposterior radiograph of the pelvis of a 38-year-old man who had MFH of bone in the acetabulum. (b) Gadolinium-enhanced axial T1-weighted MRI showing the tumor arising in the acetabulum. (c) Postoperative radiograph after reconstruction using a frozen bone autograft. (d) Follow-up radiograph 1.5 years after surgery showing osteoarthritis of the hip joint.
Consequently, the mean functional score for patients with endoprosthesis (42%) was worse than that for patients with other reconstructions (55%), which seemed to be attributed to the high complication rate in the former group. The best results were obtained in patients who underwent iliofemoral arthrodesis, although the number of patients was small. Regardless of the surgical procedures, the functional scores for patients with postoperative deep infections or dislocations were quite low; the mean MSTS scores for patients with these complications were less than 30%, whereas those for patients without these complications were more than 50%. Overall, the functional outcomes were similar between all the reconstructive options, except for those without postoperative major complications, indicating that avoiding these complications is highly important for achieving better functional outcome.
Limb-sparing surgery for periacetabular tumors is one of the most challenging procedures for orthopaedic oncologists [
Pelvic reconstruction with an endoprosthesis has been a major challenge and in this series the functional results were not satisfactory. Postoperative function was markedly affected by major complications such as deep infection or implant dislocation. To date, various reports have demonstrated the high major complication rates by endoprosthetic replacement, ranging 18–65%; 40–65% in saddle prostheses [
Patients in this series who underwent hip transposition achieved relatively good function. Hip transposition, reported in 1988 by Winkelmann at the University Hospital of Münster, is characterized by a lower incidence of complications and revision surgery in comparison with other reconstructive procedures [
Iliofemoral arthrodesis achieved the best functional scores in this series, attributable to a good gait performance with a stiff hip. Previous studies have also mentioned the advantage of a durable, pain-free, but stiff hip and less leg-length discrepancy [
Although the small number of patients who underwent reconstruction with a frozen bone autograft limits the interpretation of our results, we achieved relatively good functional results. However, the patient developed hip osteoarthritis and has complained of continuous pain. Resurfacing total hip arthroplasty may therefore be necessary in the future. As a previous report has indicated, the advantages of this procedure include a perfect fit, the lack of any need for a bone bank, easy attachment of tendons and ligaments, and a related desirable bone stock [
We found that patients with major complications had markedly reduced functional scores. Thus, reduction of the postoperative complication rate is highly important to obtain better function. In general, since the patients with pelvic sarcoma have a poor prognosis, complete resection of the tumor as well as reconstruction without postoperative complications is desirable for them. From this viewpoint, we consider that the surgical procedures without massive use of nonbiological materials, including endoprosthesis, are preferable to achieve better function and fewer complications, although the rarity and variability of these tumors preclude a statistical comparison of outcomes. Alternately, appropriate selection of reconstruction procedures for individual patients, considering the amount of remaining bone and soft tissues and novel techniques such as tissue transfer or computer-assisted surgery, would lead to fewer complications and better function.
This study summarized the clinical outcomes of major reconstructive procedures after resection of periacetabular tumors. Postoperative major complications, including deep infection or hip dislocation, remarkably worsened functional outcome. Endoprosthetic reconstruction failed without adequate soft tissue coverage. Therefore, avoiding postoperative complications is highly important for achieving better function, suggesting that surgical procedures with adequate soft tissue coverage or without the massive use of nonbiological materials are preferable. Appropriate selection of reconstruction procedures for individual patients, considering the amount of resection and remaining bone and soft tissues, would lead to fewer complications and better function.
The authors declare that there is no conflict of interests regarding the publication of this paper.
The authors would like to thank all participating patients, as well as the study nurses, coinvestigators, and colleagues who made this trial possible.