Journal of Surgical Oncology 2008;97:146–149
Clinical Outcome of Parosteal Osteosarcoma ILKYU HAN, MD, PhD,1{ JOO HAN OH, MD, PhD,3{ YEONG GON NA, MD,1{ KYUNG CHUL MOON, MD, PhD,2{ AND HAN-SOO KIM, MD, PhD1*,§ 1
Department of Orthopaedic Surgery, Seoul National University College of Medicine, Seoul, Korea 2 Department of Pathology, Seoul National University College of Medicine, Seoul, Korea Joint Reconstruction Center, Seoul National University Bundang Hospital, Seongnam-si, Gyeonggi-do, Korea
3
Background: Parosteal osteosarcoma is a rare type of osteosarcoma with distinct characteristics. Clinical outcome of 21 patients was analyzed to assess the predictive relevance of surgical margin, intramedullary tumor extension and histologic grade. Methods: There were 5 mens and 16 womens with an average age of 26 years. Average follow-up was 9.1 years (range 2.5–22.1). Most common sites were distal femur (15) and proximal humerus (2). No patient presented with metastasis. Surgical margin was wide in 13 and marginal in 8. Intramedullary extension was seen in 10 patients (48%). Focal high-grade 2 and 3 tumors were seen in 11 (52%) and 3 (14%) patients respectively. Results: Twenty patients (95%) were alive without disease. Two (10%) had relapse, one with local recurrence and another with local recurrence and lung metastasis. Of eight marginal procedures performed, 2 turned out to have histologically tumor-positive margins, both of whom later developed relapses. All patients with histologically negative margins remained disease-free. Presence of intramedullary extension and focal highgrade tumor was not significantly associated with relapse. Conclusions: A marginal but histologically negative margin of excision appears adequate for parosteal osteosarcoma. However, long-term follow-up is warranted for monitoring of rare incidences of local recurrences or distant metastases.
J. Surg. Oncol. 2008;97:146–149.
ß 2007 Wiley-Liss, Inc.
KEY WORDS: prognosis; histologic grade; surgical margin; intramedullary extension
INTRODUCTION Parosteal osteosarcoma is a rare form of osteosarcoma which usually arises on the surface of the metaphysis of long bones, representing 1–6% of all osteosarcomas [1–3]. It affects females more often than males and the peak incidence is in the third and fourth decades [4]. It is regarded as a distinct form of osteosarcoma with better prognosis as compared with conventional osteosarcoma [5]. Achieving wide surgical margins for parosteal osteosarcoma has often been difficult. Parosteal osteosarcomas tend to be large at presentation and its predilection for posterior aspect of distal femur has made achieving wide surgical margins difficult as only loose fibro-fatty tissue separates the tumor from adjacent neurovascular structures [6]. Moreover, intramedullary extension of the tumor necessitates removal of an entire segment of bone including the articular surface of the bone. The morbidity associated with wide operative procedures can be severe. Parosteal osteosarcomas may extend into the underlying cortex and penetrate into medullary cavity. The significance of intramedullary extension on oncologic outcome has been debated. Tumor extension into medullary cavity has been associated with poor oncologic outcome in some studies, as this lesion was thought to behave like high-grade conventional osteosarcomas. Others studies have shown no association between intramedullary extension and oncologic outcome [7–9]. Parosteal osteosarcoma can harbor areas of high grade tumor within predominantly low-grade lesions [10,11]. The prognostic significance of these focal high-grade lesions has been controversial. Some studies reported that patients with tumors that have these high-grade foci are at a greater risk to metastasize [12,13], whereas others showed no significant effect on the oncologic outcome [8,14]. In the present study, the clinical outcome of parosteal osteosarcoma treated at a single institution during a 21-year period is presented. The significance of the focal high-grade tumor, intramedullary extension
ß 2007 Wiley-Liss, Inc.
of the tumor and the surgical margin on the oncologic outcome in parosteal osteosarcoma was analyzed.
MATERIALS AND METHODS A retrospective review was conducted of 21 patients treated for parosteal osteosarcoma at our hospital between 1984 and 2004. The institutional review board approved this study. Patient charts were reviewed for variables including demographics, tumor site, presenting symptoms, surgical stage, details of the operation, and oncologic outcome. Histologic grade of each case was assessed by reanalyzing the pathology material by a musculoskeletal pathologist (KC Moon) according to the system by Ahuja et al. [12]. Three components (fibrous, osseous, and cartilaginous) of tumors were individually graded on the basis of cellularity, the pleomorphism and the mitotic activity. The overall grade was assigned as the highest grade assigned to any one component. Tumors were designated as grades 1, 2, or 3 for low grade, intermediate grade and high grade, respectively. In general, grade 1 tumors comprised of normochromic cells with only a few mitotic figures with little anaplastic features. Grade 2 showed cells with moderate degree of anaplasia and more mitoses than {
Assistant Professor. Resident. § Associate Professor. *Correspondence to: Han-Soo Kim, MD, PhD, Department of Orthopaedic Surgery, Seoul National University College of Medicine, 28 Yongon-dong, Chongno-gu, 110-744 Seoul, Korea. Fax: 82-2-764-2718. E-mail:
[email protected] Received 12 May 2007; Accepted 8 August 2007 DOI 10.1002/jso.20902 Published online 29 November 2007 in Wiley InterScience (www.interscience.wiley.com). {
Outcome of Parosteal Osteosarcoma
147
Fig. 1. Illustrative case of parosteal osteosarcoma treated with wide excision (Patient 12). A,B: Preoperative radiographs (A) and MRI (B) showing parosteal osteosarcoma of the distal femur (C). Postoperative radiographs after reconstruction with endoprosthesis. grade 1 tumors. Grade 3 harbored sarcomatous cells with highest degree of anaplasia and many mitotic figures. Dedifferentiation was defined as limited areas of grade 3 tumor in predominantly grade 1 lesion. For staging of the tumor according to the Musculoskeletal Tumor Society System, grade 1 lesions were assigned to stage I and grade 2 and 3 lesions to stage II. Staging workup included either MRI or CT for evaluation of local extent of the tumor and bone scan and chest CT for detection of metastasis. Because of the extracompartmental location of parosteal osteosarcomas, all the tumors were graded as either stage IB (n ¼ 7) or stage IIB (n ¼ 14) [8]. The presence of intramedullary extension of the tumor was evaluated by analyzing the pathology materials. For the cases in which pathology materials were not suitable for analysis, such as the use of neoadjuvant chemotherapy or the use of recycled autograft for reconstruction, computed tomography (CT) or magnetic resonance imaging (MRI) scans were used for the evaluation of intramedullary extension of the tumor. The presence of intramedullary extension on CT or MRI scans was determined by a musculoskeletal radiologist. Intramedullary extension was defined as extension of the tumor into the medullary cavity. Operative records and pathology materials were reviewed to evaluate the adequacy of surgical margin. A wide surgical margin procedure was recognized when a cuff of normal tissue completely encased the tumor. In the case of intramedullary extension a wide margin was given when the segment of bone with extension of the tumor was completely excised (Fig. 1). A marginal margin was recognized when a free margin of normal cortex and marrow was seen microscopically between the tumor and the bone (Fig. 2) [6]. The statistical analysis was performed using either Fischer’s exact test or Kruskal Wallis test with the SPSS for Windows statistical package (version 12.0; SPSS, Chicago, IL, 2004).
RESULTS There were 5 males and 16 females between 14 and 42 years of age (mean, 26 years) at the time of surgery (Table I). The average followup period was 9.1 years (range, 2.5–22.1 years). The most common site of the primary tumor was distal femur (15 patients), followed by proximal humerus (2), distal tibia (1), proximal femur (1), proximal fibula (1), and ilium (1). Most of the lesions were located in metadiaphyseal portion of long bones. The most common presenting symptoms were asymptomatic mass (53%) or pain (43%). Chemotherapy was not administered in most of the cases with the exception of two patients (patients 7 and 9). At the latest follow-up, 20 patients (95%) were alive without disease. One of these patients had local recurrence 10 years after index operation and has no detectable lesion at 12 years since reoperation (Patient 1). One patient underwent re-resection of the tumor for local recurrence 3 years after index operation (Patient 8). This patient developed pulmonary metastases 4 years after reoperation and died of the disease. Three of 11 patients with prostheses had revision surgeries for complications of deep infection (1), prosthesis fracture (1) and loosening (1). No patients reconstructed with recycled autograft or bone cements had complications that necessitated additional operations. One patient suffered fracture of the fibula graft used for arthrodesis, which was treated successfully with internal fixation. Overall, ten cases (48%) showed evidence of intramedullary extension of the tumor. Of the 17 cases in which pathology material was used for the evaluation, seven (41%) had intramedullary extension of the tumor. Of the four cases where no suitable pathology material was available, three (75%) showed evidence of intramedullary extension on CT or MRI scans. Two (20%) out of ten cases with intramedullary extension of the tumor had tumor relapse. One (patient 1) of these patients had local recurrence and another
Fig. 2. Illustrative case of parosteal osteosarcoma treated with marginal excision (Patient 17). A,B: Preoperative radiographs (A) and MRI (B) showing parosteal osteosarcoma of the distal femur (C). Postoperative radiographs after reconstruction with cementing. Journal of Surgical Oncology DOI 10.1002/jso
Han et al.
Marginal (þ) Wide () Marginal () Wide () Wide () Wide () Wide () Marginal (þ) Wide () Wide () Wide () Wide () Marginal () Wide () Wide () Marginal () Marginal () Marginal () Marginal () Wide () Wide () No Amputation Prosthesis Prosthesis Arthrodesis Prosthesis Prosthesis No Prosthesis Prosthesis Prosthesis Prosthesis Prosthesis Prosthesis Recycled autograft Cementing Cementing Recycled autograft No Arthrodesis Prosthesis
271 102 38 229 219 211 184 78 180 31 92 87 67 66 43 33 30 33 91 132 52
Yes (122) No No No No No No Yes (38) No No No No No No No No No No No No No
No No No No No No No Yes (78) No No No No No No No No No No No No No
NED NED NED NED NED NED NED DOD NED NED NED NED NED NED NED NED NED NED NED NED NED
(patient 8) had local recurrence and distant metastasis. Eleven patients without intramedullary extension remained disease-free. There was no statistically significant association between intramedullary extension and local recurrence or distant metastasis (P > 0.05). Most of the patients in the early part of this series were managed with a wide excision. We began to perform marginal excision for cases in which the articular surface could be spared so that a better functional outcome or less morbidity could be expected. Overall, resection of the entire segment of the bone with the intention to achieve wide operative margin was performed in 15 cases. After pathological review of resected specimens, two (Patients 3 and 13) of the 15 cases were determined to have marginal margin. Six cases underwent resection of the tumor with a marginal margin. Of the 8 marginal procedures performed, two turned out to have histologically tumor-positive surgical margins (patients 1 and 8), which was considered as tumorcontaminated intralesional surgical margins. Both of them later developed local recurrences, which was described previously. All the patients with histologically negative surgical margins remained disease-free. The association between histological margin and disease relapse was statistically significant (P ¼ 0.01). Seven of the 21 patients (33%) were identified as having histological grade 1 tumors. Although the remaining 14 patients had predominantly histological grade 1 disease, eleven (52%) and three (14%) displayed focal areas of grade 2 and 3 tumors respectively, and were designated as such. The two patients with tumor recurrence each had grade 2 tumors. The three patients with grade 3 tumors remained disease-free at last follow-up. There was no significant association between histologic grade and disease relapse (P > 0.05).
DISCUSSION
Journal of Surgical Oncology DOI 10.1002/jso
NED, no evidence of disease; ALW, alive with disease; DOD, dead of disease. a Time to local recurrence, metastasis and death (months).
23/F 23/F 16/F 28/F 34/M 14/M 16/F 35/F 15/M 26/F 42/M 36/F 18/F 32/F 17/F 32/F 22/F 22/M 35/F 43/F 16/F 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Dist. femur Dist. femur Dist. femur Dist. femur Prox. humerus Dist. femur Dist. femur Dist. femur Dist. tibia Dist. femur Prox. femur Dist. femur Dist. femur Dist. femur Prox. fibula Ilium Dist. femur Prox. humerus Dist. femur Dist. femur Dist. femur
2 1 1 2 2 2 1 2 3 2 1 2 1 2 2 3 2 3 1 2 1
Mass Pain Mass Mass Mass Mass Pain Incidental Pain Pain mass Pain Mass Pain Mass Mass Pain mass Pain Mass Mass Mass Pain Mass
IIB IB IB IIB IIB IIB IB IIB IIB IIB IB IIB IB IIB IIB IIB IIB IIB IB IIB IB
Yes Yes No No Yes Yes Yes Yes Yes No No Yes Yes Yes No No No No No No No
Resection margin Symptom Histologic grade Location Age/Sex Case
TABLE I. Details of the 21 Patients With Parosteal Osteosarcoma
Stage
Intramedullary involvement
Reconstruction
FU (mos)
Local recurrence (mos)a
Metastasis (mos)a
Onocologic status (mos)a
148
As parosteal osteosarcomas are usually located in the metaphysis of long bones, removal of the articular surface of the bone is often required to achieve a wide operative margin. Moreover, intramedullary extension of the tumor may necessitate removal of the entire segment of bone [15]. The frequent extracompartmental location in the popliteal fossa or axilla makes achieving wide margin difficult without sacrificing the neurovascular bundle. The morbidity of wide operative procedure does not seem to be justified, as good local disease control is possible with marginal resections [8]. In our series, marginal but histologically negative surgical margins were adequate for preventing local recurrence and distant metastasis, which was comparable with the results of hemicortical resection for local control of the tumor by Lewis et al. [15]. The histological prevalence (48%) of intramedullary extension of the tumor was similar to those of previous reports [6–8]. However, when CT or MRI was used for the evaluation, seventeen cases (81%) showed evidence of intramedullary invasion. The disparity may be partly due to the improvement of radiologic imaging tools, most noticeably magnetic resonance imaging. Careful interpretation of these findings on MRI is needed in planning of the treatment. Tumor extension into the medullary canal has been reported as an adverse prognostic factor in some studies while others studies showed no difference in the outcome. Of the 17 cases with adequate pathology specimens, two (29%) of the seven patients with intramedullary involvement had tumor relapse while none of the 10 patients without medullary involvement had relapse. However, the number of cases without intramedullary extension was too small to show a significant effect of intramedullary extension on oncologic outcome. There were three cases (14%) in this series which had foci of grade 3 histology lesions that qualified as dedifferentiated type of parosteal sarcoma. No relapses were seen in these cases, which is contradictory to some studies that reported the increased risk of relapse in cases with high-grade foci [12]. The relapses in parosteal osteosarcoma can be seen in many years after the operation [8], which was evident in this
Outcome of Parosteal Osteosarcoma series of patients. Two out of the three cases had a relatively short 33 months of follow-up, which may have contributed to this finding. We now perform a marginal excision when a better functional outcome or less morbidity can be anticipated compared with a wide procedure. In cases located in the posterior part of the distal femur, which is the most frequent location of parosteal osteosarcoma, we perform a marginal excision if the articular surface of the distal femur can be saved. Marginal excision usually results in hemicortical defect of the bone, which we reconstruct with cement or allograft. Relatively small number of patients limits the interpretation of the results of this study. Prospective comparison in a larger number of patients with a longer follow-up is necessary to determine the prognostic significance of focal high-grade tumor, intramedullary extension and surgical margin on oncologic outcome in parosteal osteosarcoma. In summary, marginal but histologically negative surgical margin seems adequate for local control of parosteal osteosarcoma. However, long-term follow-up is warranted for monitoring of rare incidences of local recurrences or distant metastases.
REFERENCES 1. Geschickter CF, Copeland MM: Parosteal osteoma of bone: A new entity. Ann Surg 1951;133:790–807. 2. Heul V, Ronnen R: Juxtacortical osteosarcoma; Diagnosis, differential diagnosis, treatment and an analysis of eighty cases. J Bone Joint Surg 1967;49-A:415–439. 3. Dwinnell LA, Dahlin DC, Ghormley RK: Parosteal (juxtacortical) osteogenic sarcoma. J Bone Joint Surg Am 1954;36-A:732–744.
Journal of Surgical Oncology DOI 10.1002/jso
149
4. Enneking W, Springfield D, Gross M: The surgical treatment of parosteal osteosarcoma in long bones. J Bone Joint Surg 1985;67A:125–135. 5. Ritschl P, Wurnig C, Lechner G, et al.: Parostela osteosarcoma 2– 23-year follow-up of 33 patients. Acta Orthop Scand 1991;62: 195–200. 6. Campanacci M, Picci P, Gherlinzoni F, et al.: Parosteal osteosarcoma. J Bone Joint Surg 1984;66-B:313–321. 7. Kavanagh T, Cannon S, Pringle J, et al.: Parosteal osteosarcoma. Treatment by wide local resection and prosthetic replacement. J Bone Joint Surg 1990;72-B:959–965. 8. Temple H, Scully S, O’Keefe R: Clinical outcome of 38 patients with juxtacortical osteosarcoma. Clin Orthop 1999;373:208– 217. 9. Wold LE, Unni KK, Beabout JW, et al.: Dedifferentiated parosteal osteosarcoma. J Bone Joint Surg Am 1984;66:53–59. 10. Okada K, Frassica F, Sim F: Parosteal osteosarcoma: A clinicopathologic study. J Bone Joint Surg 1994;76-A:366–378. 11. Scaglietti O, Calandriello B: Ossifying parosteal sarcoma. Parosteal osteoma or juxtacortical osteogenic sarcoma. J Bone Joint Surg Am 1962;44-A 635–647. 12. Ahuja S, Villacin A, Smith J, et al.: Juxtacortical (parosteal) osteogenic sarcoma. Histological grading and prognosis. J Bone Joint Surg 1977;59-A:632–647. 13. Unni KK, Dahlin DC, Beabout JW, et al.: Parosteal osteogenic sarcoma. Cancer 1976;37:2466–2475. 14. Raymond AK: Surface osteosarcoma. Clin Orthop Relat Res 1991;270:140–148. 15. Lewis V, Gebhardt M, Springfield E: Parosteal osteosarcoma of the posterior aspect of the distal part of the femur. J Bone Joint Surg 2000;82-A:1083–1088.
