Periosteal osteosarcoma

A single-institution experience




Periosteal osteosarcoma is a rare variant of osteosarcoma. Wide surgical removal is the mainstay of treatment, but controversy remains about the role of chemotherapy. The objective of this study was to review and analyze the clinical and treatment-related factors that influence the survival of patients with periosteal osteosarcoma who received treatment in a single institution.


Thirty-three patients with periosteal osteosarcoma (19 males and 14 females) with a median age of 16 years (range, ages 6-32 years) underwent surgery (32 patients) and received radiotherapy (1 patient). Chemotherapy was received according to different regimens for high-grade osteosarcoma by 14 patients who had grade 3 tumors.


The 10-year overall survival rate was 84%. The only patient who did not undergo surgery died of disease after 9 months; for the remaining 32 patients the 10-year disease-free survival rate was 65%. Survival was not influenced by the receipt of chemotherapy. The patients who received chemotherapy had a 10-year overall survival rate of 86%, and those who received only local treatment had an overall survival rate of 83% (P = .73).


The authors' experience indicated that the treatment of periosteal osteosarcoma requires only wide surgical removal of the tumor and that adjuvant chemotherapy does not improve survival. Cancer 2011. © 2010 American Cancer Society.

Periosteal osteosarcoma is a rare tumor that represents <2% of all osteosarcomas. It originates from the periosteum, usually in the diaphysis of long bones. The histologic pattern of periosteal osteosarcoma is characterized by large, irregular lobules of chondroblastic tissue with small areas of osteoid formation.1-6 The typical radiologic features are periosteal, fusiform, well defined masses encircling the bone and thickening the cortex with scalloping and perpendicular periosteal reaction extending into a broad-based soft tissue mass.1-7

Periosteal osteosarcoma is an intermediate-grade chondroblastic osteosarcoma arising on the surface of bone. Periosteal osteosarcoma, as an intermediate tumor, has less aggressive biologic behavior with a lower propensity to metastasize compared with conventional osteosarcoma or high-grade surface osteosarcoma.4-6, 8-11 Conventional osteosarcoma is a primary, intramedullary, high-grade malignant tumor in which neoplastic cells produce osteoid, even if only in small amounts. Osteosarcoma is the most common nonhemopoietic, primary malignant tumors of bone with an estimated incidence of 4 to5 per million population. High-grade surface osteosarcoma is a high-grade, bone-forming malignancy that arises from the surface of bone. High-grade surface osteosarcoma is a rare entity and comprises <1% of all osteosarcomas. Conventional osteosarcoma and high-grade surface osteosarcoma share the same prognosis and the same treatment. The use of surgery is curative in only <15% of patients, whereas the addition of systemic chemotherapy has resulted in survival rates of 60% to 80% for those patients who do not have clinically evident metastases at presentation. Another osteosarcoma arising on the bone surface is parosteal osteosarcoma, which is a low-grade osteosarcoma that accounts for approximately 4% of all osteosarcomas. Its prognosis is excellent, with a 91% overall survival (OS) rate at 5 years with surgical treatment alone, and it does not require the use of chemotherapy.4, 9-11 Wide excision of the tumor is the mainstay of therapy for periosteal osteosarcoma, although controversy remains about the role of chemotherapy.8, 12-15 The objective of the current, retrospective study was to analyze the clinical and treatment-related factors that influenced the outcome of patients with periosteal osteosarcoma who underwent treatment at a single institution with special emphasis on the role of chemotherapy.


Patients with diagnosis of periosteal osteosarcoma were identified from the database of our Bone Tumor Center. Clinical data were collected from clinical records. Completeness of the data regarding types of local and systemic treatment, radiographic imaging studies, and histologic material had to be available to include patients in the study. A radiologic and histologic revision was carried out for all patients. According to the cytologic atypia of the cartilaginous component, tumors were graded in 3 tiers.

Survival analysis was performed by using Kaplan-Meier survival curves that were compared using the log-rank test. OS was calculated from the date of diagnosis at our institution to the most recent follow-up evaluation or death; event-free survival (EFS) was calculated from the date of admission to our institution to the last follow-up, recurrence, toxic death, or secondary malignancy and disease-free survival (DFS) was calculated from the date of complete remission to recurrence or last follow-up with no sign of disease.

In total, 39 patients with periosteal osteosarcoma were identified. Five patients were not included because they had insufficient clinical information (1 patient was lost to follow-up after surgery), radiologic information (in 4 patients, the radiographs were not available), and pathologic information (in 3 patients, histologic slides for revision were not available). One patient was classified with a high-grade surface osteosarcoma and was excluded based on a predominant osteoblastic and fibroblastic differentiation pattern and high-grade cytologic atypia. Overall, 33 patients were included in the current analysis.


The median age of patients at the time of admission to our institution was 16 years (range, 6-32 years), 19 patients (58%) were male, and 14 patients (42%) female. The most common primary tumor sites were the tibia (16 patients) and femur (15 patients) followed by the humerus (2 patients).

Twenty-one patients (64%) were diagnosed first at our institution. Twelve patients (36%) patients were admitted for disease recurrence or local disease progression disease after treatment that they received elsewhere, including 8 patients who underwent surgical excision 2 to 12 months before admission, 2 patients who underwent resection 2 months and 36 months before admission, and 2 patients received previously radiotherapy only.

All 33 patients had localized disease at the time of admission to our hospital. Serum alkaline phosphatase (SAP) levels were available for 20 patients, and lactate dehydrogenase (LDH) levels for were available for 14 patients. Five patients (25%) had high SAP values (25%), and 4 patients (29%) had elevated LDH values. The extent of medullary involvement was assessable in 23 patients, and only 7 patients (30%) did not have medullary invasion detected. The histologic tumor grade was grade 2 in 2 patients (6%) and grade 3 in 31 patients (94%). Thirty-two patients (97%) underwent surgery, and 1 patient who had a large tumor that was localized to the femur refused the planned surgical treatment (amputation) and was treated with radiotherapy alone. In 8 patients (25%), the local treatment was amputation; and 24 patients (75%) underwent resection. Surgical margins defined according to Enneking et al16 were available for 17 patients: Fourteen patients achieved wide surgical margins, 2 patients had margins that were categorized as marginal, and 1 patient had contaminated margins.

Six patients were admitted to hospital in the prechemotherapy era (before 1973), and 14 patients (42%) received chemotherapy (all had grade 3 tumors). Chemotherapy was delivered according to the regimens used for classic osteosarcoma (Table 1). Four patients received neoadjuvant treatment, and, in 3 patients, the percentages of chemotherapy-induced necrosis evaluated in surgical specimen was 70%, 85%, and 95%. Nineteen patients (58%) patients received no systemic treatment.

Table 1. Chemotherapy Protocols for Patients With Periosteal Osteosarcoma
PatientPrimary SiteAge, yChemotherapy RegimenTypeSurgical Complete RemissionStatus at Last Follow-Up
  1. ADM indicates doxyrubicin; CDP, cisplatin; IFO, ifosfamide; MTX, methotrexate; VCR: vincristine; NED II, no evidence of disease after relapse; VCRS, vincristine, NED, noevidence of disease; DOOC, died of other cause; ia, intra-arterial administration; DOD, died of disease.

1Femur16ADM-CDP-IFO MTXAdjuvantYesNED II, 111 mo
2Tibia15ADM-MTX-VCRAdjuvantYesNED, 215 mo
3Femur30ADM-CDP-IFO MTXAdjuvantYesDOOC, 128 mo
4Tibia15UnknownAdjuvantYesNED, 183 mo
5Tibia12ADM-CDP-IFO MTXNeoadjuvantYesNED, 73 mo
6Femur16MTX-CDP ia, hypertaermiaNeoadjuvantYesDOD, 20 mo
7Femur17ADM-CDP ia-MTXNeoadjuvantYesNED, 122 mo
8Femur17ADM-CDP-IFO MTXAdjuvantYesNED, 41 mo
9Tibia32ADM-CDP-IFOAdjuvantYesNED II, 189 mo
10Tibia13ADM-MTX-VCRAdjuvantYesDOD, 20 mo
11Femur15UnknownNeoadjuvantYesNED, 195 mo
12Femur22UnknownAdjuvantYesNED II, 109 mo
13Femur17ADMAdjuvantYesNED, 158 mo
14Tibia24ADM-CDP-IFO MTXAdjuvantYesNED, 62 mo

At a median observation period of 121 months (range, 9-367 months), 27 patients (82%) remained alive. Four (12%) patients died of disease between 9 months and 20 months after diagnosis (median, 18 months). Two patients (6%) died of mesenchymal chondrosarcoma and heart failure 128 months and 44 months after diagnosis. One patient (3%) remained alive with disease 29 months after diagnosis. At a median follow-up of 138 months (range, 18-367 months), 26 patients (79%) remained alive with no evidence of disease, including 19 patients (58%) who were continuously disease-free and 7 patients (21%) who underwent further surgical remission after disease recurrence.

Eleven patients (33%) had recurrent disease after a median time to recurrence of 14 months (range, 3-32 months). The pattern of relapse was local in 7 patients (21%), lung metastases occurred in 3 patients (9%), and 1 patient developed multiple metastases (lung, bones, and soft tissues of the contralateral tibia). Three patients had inadequate surgical margins, and 2 of those patients developed local recurrences. All but 1 patient who developed recurrent disease underwent further surgery, and, in 2 patients, chemotherapy was added. Second malignant neoplasms were observed in 3 patients (9%). Mesenchymal chondrosarcoma of the retroperitoneum developed in 1 patient 7 years after undergoing combined surgery and chemotherapy, and the patient died of the secondary malignancy. Acute lymphoid leukemia and breast cancer were detected in 2 patients who currently remain alive without disease. The diagnosis of acute lymphoid leukemia was made 2 years after the start of treatment in 1 patient who underwent surgery and received chemotherapy. Breast cancer was diagnosed 10 years after surgical treatment.

The 10-year OS rate was 84%. The only patient who did not achieve complete surgical remission died of disease 9 months after diagnosis. The 10-year EFS rate was 52%. For the 32 patients who achieved complete surgical remission, the DFS probability at 10 years was 65%. According to the medullary extent of disease, the 10-year DFS probability was 61% when medullary invasion was present and 86% when it was absent (P = .25), and the 10-year OS rates were 74% and 86%, respectively (P = .73). Tumor recurrence (2 local recurrences and 4 distant metastases) was observed in 6 of the 16 patients (37.5%) who had medullary invasion. One (14%) local recurrence was observed in the 7 patients who did not have medullary involvement. Medullary extension of the tumor was detected in 16 patients (69%). Nine patients received chemotherapy only, whereas 7 patients underwent surgery only. In the chemotherapy group, 1 patient died of a second malignant tumor (mesenchymal chondrosarcoma of the retroperitoneum), 4 patients had distant metastases, and 4 patients were continuously disease-free. Two patients with metastases underwent surgery and achieved a second complete surgical remission. Two patients died of disease. In the surgical group, 5 patients were continuously disease-free, and 2 patients developed local recurrences. After surgery for the local recurrence, those 2 patients were alive in second complete remission. The 2 patients who had grade 2 tumors remained alive and disease-free after a minimum observation time of 5 years. The 10-year OS and DFS rates were 83% and 66%, respectively, for patients with grade 3 periosteal osteosarcoma.

The patients who received chemotherapy had a 10-year OS rate of 86%, and those who received local treatment only had a 10-year OS rate of 83% (P = .73) (Fig. 1). The 10-year EFS was 47% for the chemotherapy group and 66% for those who did not receive chemotherapy (P = .25). The 10-year DFS rate was 61% for the chemotherapy group and 86% for those who did not receive chemotherapy (P = .25). The 10-year OS rate was 67% for patients who had pulmonary metastases and 86% for patients who relapsed with a local recurrence. The patient who relapsed with multiple metastases died 4 months after the relapse.

Figure 1.

Chemotherapy treatment and cumulative (Cum.) overall survival are illustrated.


The current study, to our knowledge, is the largest single-center series of patients with periosteal osteosarcoma to date. The survival data were consistent with those reported by other authors, with a 10-year rate OS of 84%, indicating that patients with this tumor have a better prognosis than patients with high-grade surface osteosarcoma and a worse prognosis than patients with parosteal osteosarcoma.9-11

All patients in our study developed recurrent disease within the first 3 years after diagnosis. It is noteworthy that, of the 12 patients who relapsed, 7 patients had a local recurrence, and 5 patients had distant metastases. The pattern of recurrence in series that reported data on classic osteosarcoma was different, in that most patients had distant rather than local recurrences.17 Furthermore, the postrelapse survival probability in patients with periosteal osteosarcoma patients was high, and the results were better for the patients who had local recurrences than for those who had pulmonary metastases (86% vs 67%). In patients with classic, high-grade osteosarcoma, the reported probability of postrelapse survival is <30% at 5 years.18, 19

The current data confirm that this variant of osteosarcoma has less aggressive behavior compared with that of classic high-grade osteosarcoma, including a lower propensity for systemic recurrence. Periosteal osteosarcoma is a bone surface tumor and, according to some authors, medullary invasion should exclude the diagnosis of periosteal osteosarcoma.2 It is likely that this condition reflects tumor progression and its biologic behavior. In our series, medullary involvement was assessable in 23 patients; and, in 70% of those patients, medullary invasion was detected. These data clearly demonstrated that intramedullary extension is common in patients with periosteal osteosarcoma and does not distinguish a different pathologic entity.

In our series, this condition was related to a greater probability of tumor recurrence that, nevertheless, did not lead to a lower probability of survival. It is noteworthy that the local recurrence rate was similar in the 2 groups (local recurrences were observed in 1 of 7 patients with no medullary invasion and in 2 of 16 patients with medullary extension of the tumor), whereas distant metastases were detected only in patients who had medullary involvement. This finding suggests that medullary invasion may identify patients who have more aggressive tumors. However, 2 of the 4 patients who had metastases were long-term postrelapse survivors. The low or intermediate neoplastic activity of this tumor is a possible explanation for this apparent discrepancy between the recurrence rate and OS.

The role of chemotherapy in the treatment of periosteal osteosarcoma is controversial. The largest study ever reported in literature is the European Musculoskeletal Oncology Society study, in which a survey was carried out on the treatment of this disease. It is worth noting that >66% of the 119 patients (81 patients) whose data were collected received some chemotherapy treatment. In that study, no differences in survival were observed between patients who did or did not receive chemotherapy. That study had many limitations; first, a central revision of the histology and imaging was not performed, and tumor grade was not reported for 43% of the patients. In our series, all patients who were included in the analysis underwent histologic and radiologic revision. Only patients with grade 3 tumors received chemotherapy; and, also in our study, the use of chemotherapy was not beneficial for patients who had periosteal osteosarcoma.

The indication to administer chemotherapy was influenced over the years by uncertainty regarding the possible advantage offered by its use in this particular histotype. Starting from late 1970s, chemotherapy always was proposed before surgery for patients who had large tumors. The decision to use chemotherapy in an adjuvant setting was always shared with patients. Only a randomized study properly may answer the question of whether or not chemotherapy is beneficial for patients with periosteal osteosarcoma. Unfortunately, the rarity of this disease does not permit such a study. Our data on chemotherapy for patients with a higher probability of recurrence (ie, those with medullary involvement) do not seem support its use even in this group of periosteal osteosarcomas. In fact, metastases developed in 4 of the 9 patients in our study who underwent combined surgery and chemotherapy; whereas, in the surgical group, 2 patients developed only local recurrences.

Data from the European Musculoskeletal Oncology Society survey and from the current study suggest that patients with periosteal osteosarcoma should not receive chemotherapy. Three second malignancies were discovered in our series (9% of patients). Two of those patients had received previous chemotherapy. The percentage of second malignant neoplasms observed in the current study was higher than the percentage reported among long-term survivors of classic osteosarcoma.17, 20 Unfortunately, no cytogenetics were available from those studies, and the low numbers did not allow those investigators to identify chemotherapy as a possible cause of second malignancy.

In conclusion, patients who have periosteal osteosarcoma have a better probability of survival than patients who have classic osteosarcoma. The key factor for treatment is wide surgical removal of the tumor. Frontline chemotherapy does not improve the probability of survival for these patients.


We thank Cristina Ghinelli, medical artist, for the graphics.


The authors made no disclosures.