An increasing number of patients are surviving their primary cancers.16,22 Unfortunately, the prevalence of secondary cancers is also increasing.16,22 Although some cancer diagnoses may be associated with the development of a secondary cancer, some arise as a result of the treatment of their primary cancer. These secondary cancers have been termed iatrogenic diseases of success.23
The carcinogenic effects of ionizing radiation are well known.18,19,27 Prior radiation of bone is a well-defined risk factor for the development of secondary sarcomas.6,16,22 In 1948, Cahan et al proposed four criteria for the diagnosis of postradiation sarcoma: (1) history of radiation to the site; (2) histologically proven malignant tumor arising within the radiated field; (3) a 3 to 4 year latency period between radiation and development of the sarcoma; and(4) an original histology different from the new tumor.10 Osteosarcoma is the most common postradiation sarcoma of bone.13 Postradiation osteosarcoma was originally associated with a poor prognosis.16,18,20,24 Postradiation osteosarcoma patients are frequently older patients with more comorbidities.15 Most patients died of metastases and failure of local control.18,21,24 Although chemo- therapy was often used, it was not thought to be effective. The substantial comorbidities of older patients can often prevent full dose chemotherapy regimens; however, even in the young patient population (less than 24 years of age) without the associated comorbidities, chemotherapy may be ineffective.15,21,22 Many authors postulated more intensive chemotherapy regimens may reduce recurrence and metastatic rates.
Local control of osteosarcoma primarily is a surgical problem, but histologic response to neoadjuvant therapy is the single most important prognostic factor for long-term survival.5,12,28 Several studies have shown the degree of necrosis induced by chemotherapy correlates with disease- free survival. Good responders, those patients with > 90% necrosis, have less probability of developing metastasis, local tumor recurrence or death.2,3,17,30 In studies reporting the outcome of postradiation osteosarcoma in the era of contemporary chemotherapy, however, only overall survival and not histologic response to neoadjuvant chemotherapy was examined. We hypothesized if current chemotherapy regimens improve survival of postradiation osteosarcoma patients, this should be reflected in the histologic response of the resected specimen to the chemo- therapy. Histologic response to neoadjuvant therapy should correlate with survival. However, not all patients with a favorable response to chemotherapy survive.
Since modern chemotherapy regimens have substantially improved the outcome of patients with primary extremity osteosarcoma,4,25,28-30 the question arises if, as many authors have postulated, these more aggressive and intensive protocols have resulted in reduced recurrence and metastatic rates and survival for patients with postradiation osteosarcoma. There have been conflicting reports on the overall prognosis and survival and, recently, some authors have suggested the prognosis may not be as grave if contemporary chemotherapy regimens are administered.6,7,26 Therefore, it is not clear whether the outcome of postradiation osteosarcoma has improved in the era of contemporary multi-agent chemotherapy.
We hypothesized that despite modern chemotherapy protocols the prognosis of patients with postradiation chemotherapy remains grave. We asked whether the histologic response to chemotherapy and the latency period from administration to development of the postradiation osteosarcoma would predict disease free or overall survival.
MATERIALS AND METHODS
We retrospectively reviewed our departmental database and the institutional tumor registry to identify 27 patients diagnosed and treated for skeletal postradiation osteosarcoma from 1980-2003. The histologic diagnosis of osteosarcoma was confirmed by one pathologist (AKR). Patients with other postradiation sarcomas were excluded from the study. Of the 27 patients, 19 were female and 8 male. The median age at diagnosis was 54 years (range,12-86 years). Twenty-six patients presented with localized disease, and one patient presented with lung metastases. The mean age at diagnosis of the first cancer was 33 years (range, 2-67 years). The most frequent malignancy was breast cancer (9 patients), followed by cervical cancer (4 patients) soft tissue sarcomas (4 patients), Ewing's sarcoma (2 patients), Hodgkin's lymphoma (2 patients), sarcoma of bone (2 patients), thyroid cancer (1), mediastinal seminoma (1), giant cell tumor of bone (1) and prostate cancer (1) (Table 1). The anatomic sites of osteo-sarcoma were the pelvis (5 patients), femur (5 patients), scapula (5 patients), sternum/chest wall (5 patients), humerus (3 patients), clavicle (2 patients), and tibia (2 patients) (Table 3). The average time to the development of the osteosarcoma (latency) was 18.3 years (range, 5-38 years). Our study was approved by the institutional review board of the University of Texas, MD Anderson Cancer Center.
Nine patients received disease-specific chemotherapy for their primary diagnosis. All patients received radiation. The mean dose of radiation was 50.1 Gy (range, 30-65 Gy). The exact dose of radiation was unavailable for seven patients.
Medical records were reviewed to identify the latency period from the time of radiation to the development of postradiation osteosarcoma, the anatomic site, metastatic disease, surgical procedure, type of adjuvant treatment, and response to chemo- therapy as measured by percent necrosis. Overall survival was calculated from the day of osteosarcoma diagnosis until death from any cause.
Twenty-two of the 27 patients received induction chemo- therapy for their osteosarcoma. The patients received the chemotherapy regimens according to the MD Anderson protocols for osteosarcomas used at the time of their diagnosis. The basic neoadjuvant chemotherapy regimen for osteosarcomas was Adriamycin (Bedford Laboratories, Bedford, OH) (90 mg/m2) and intra-arterial cisplatin (120 mg/m2). However, doses were adjusted appropriately for age, performance status, and chemo- therapy history (Table 2).
Twenty-one patients underwent limb sparing surgery and six patients underwent amputation. Margins were negative in all but one patient. The entire radiation field was not routinely resected. Eleven patients received postoperative chemotherapy (Table 2). All specimens were examined to determine the chemotherapy- induced necrosis rate. The average necrosis rate was 63.5% (range, 12-100%), but seven patients had ≥ 90% necrosis. Four of these patients with ≥ 90% necrosis died of their disease at an average of 16.5 months (range, 7-22 months).
Survival probabilities were calculated using the Kaplan-Meier method. Differences in overall survival time as a function of various prognostic factors were examined by a log rank test. Two sided p values were reported (p < 0.05). Multivariate analysis was performed using a Cox proportional hazard model.
Survival remains grave. The 5-year Kaplan-Meier survival estimate was 27.2% (95% confidence interval [CI], range76.3-99.8%) (Fig 1). Nineteen patients died of their disease at a mean of 26 months (range, 2-119 months), and one patient died of chemotherapy related causes. Seven of the 27 patients (26%) are alive with no evidence of disease at a median followup of 92.8 months (range, 32-218 months). Patients developed pulmonary metastases as their primary site of metastatic disease. The necrosis rate of these patients ranged from 32-95%. Of the seven patients with ≥ 90% necrosis only three are alive at a median followup of 117.3 months (range, 32-218). Three of the 27 patients developed recurrent disease. One patient developed recurrent soft tissue disease; the soft tissue mass was resected and the patient is now disease free. One patient developed a local intra-osseous recurrence, underwent amputation, and is now disease free. One patient developed a local soft tissue paravertebral recurrence within the radiated site and died of metastatic disease. Local failure did not impact overall survival. One patient was lost to followup at 3 months.
Histologic response of the chemotherapy of the resected osteosarcoma did not predict disease free or overall patient survival. We found no difference for survival based on patient age at diagnosis of the first cancer, age at diagnosis of the postradiation osteosarcoma, gender, primary cancer, whether chemotherapy was given for the original diagnosis, radiation dose, osteosarcoma site, presence of metastases at diagnosis, surgical procedure, margin, type of adjuvant treatment, or chemotherapy-induced necrosis.
The latency period from the administration of radiation to the development of the postradiation osteosarcoma correlated (p < 0.0016) with survival (Fig 2). Patients who had a latency of greater than 10 years had a better prognosis (p < 0.0016). When the latency period from radiation treatment to the diagnosis of the osteosarcoma was included in the Cox proportional hazard model, we identified no other predictors for survival.
Historically, the prognosis of patients with postradiation osteosarcoma has been grave. Many of the studies supporting this observation were based on patients who were treated before the development of effective multi-agent chemotherapy. The data suggested more aggressive and intensive chemotherapy regimen may result in reduced recurrence and metastatic rates for patients with postradiation osteosarcoma.15,16,18,20,24 There have been some reports on the outcome of postradiation osteosarcoma in the era of multi-agent chemotherapy, however, the results have been conflicting.
We note several limitations. One of the drawbacks of retrospective studies is that the treatments already have been administered and often, as in this study, vary. However, even though the patients in this study did not all receive the same regimen, the regimens given were the accepted standards at the time of diagnosis and based on the same chemotherapeutic agents. Another treatment variable of our study is the fact that the doses were adjusted age and performance status. Our series contained both pediatric and adult patients. Some patients received a more aggressive dosing because of their improved performance status and lack of comorbidities. Because of the demographics and nature of secondary osteosarcoma, patients may be limited in their treatment options given the chemotherapy administered for their primary cancer and their intolerance of standard chemotherapy regimens because of their health status. Standardization of chemo- therapy regimens in the future for this patient population is unlikely.
Our data suggest that despite multi-agent chemo- therapy, postradiation osteosarcoma continues to have a poor prognosis. The overall patient survival is less for these tumors than for conventional osteosarcoma treated with the same cytotoxic agents.28,29 Patients with secondary osteosarcomas, specifically postradiation osteosarcoma, are usually older than patients with primary osteosarcoma.16,20 The majority of the patients in this series were older than 45 years.
It has been postulated that the advanced age and the associated comorbidities of these patients may make their medical management more difficult.13,15 This is reflected in our data. Although these patients received aggressive preoperative chemotherapeutic regimens, their altered performance status may have also played a part in the modifications of the postoperative regimen which may have impacted their survival.
Several studies have shown the degree of necrosis induced by chemotherapy correlates with disease free survival. These studies suggest patients with greater than 90% necrosis have less probability of metastasis, local recurrence, or death.2,3,17,30 In our study, the necrosis did not correlate with overall survival. The reason behind this is unclear. Patient age, radiation dose, osteosarcoma site, presence of metastases at diagnosis, chemotherapy, surgical procedure, margin, and latency period did not distinguish these patients. It may well be inherent in the biology of the postradiation osteosarcoma. Although morphologically similar, its behavior is not similar to conventional primary osteosarcoma.
In our series, the poor overall patient survival was consistent with reports of others for postradiation osteosarcoma performed before and during the era of contemporary chemotherapy. Hamre et al evaluated 133 patients with secondary osteosarcomas and reported a 5-year survival rate of 17%.14 They did not find a difference in overall survival between secondary osteosarcomas occurring within irradiated fields and those that did not.8 Although Bielack et al examined 30 patients with secondary sarcomas and reported 7-year actuarial overall survival of 50% for all patients, the authors reported a 39% survival rate for those patients with postradiation osteosarcoma.6 However, Tabone et al reviewed the experience at the Société Française d'Oncologie Pédiatrique and reported a slightly better overall survival probability of 50% at 8 years.26 The average age of patients with postradiation osteosarcoma, in these studies, was younger than our patients. The younger age and better performance status may have contributed to a better prognosis. However, it should be noted the survival reported by Tabone and colleagues is still lower than the survival reported for primary nonmeta- static osteosarcoma.26,28
In the initial reports of postradiation osteosarcoma, it was noted the preexisting tumor for which the radiation therapy had been given was osseous (ie, giant cell tumor of bone and Ewing's).1 This may have accounted for the higher male to female ratio that was originally reported by Cahan.10 However, our study, like that of more recent studies, reported a sex ratio of 2.5:1, females to males.
Frassica and coauthors attributed the female predominance to the use of ionizing radiation for the common cancers of women such as breast and cervix.13 This is supported by our data, since the primary disease of 13 of the patients who received radiation was carcinoma of the breast or cervix. As randomized clinical trials mature and as national treatment guidelines become more established, the utilization of radiation therapy in the treatment of breast cancer has increased.9,11 It is important that women receive counseling about the treatment options, including ramifications and potential risks. The development of postradiation sarcoma should be included in the discussion of the potential risks of the radiation therapy. It is an uncommon but potentially fatal complication.
In most series, the median latency period ranged from 8 to 16 years.16,18,20,24,26 In our study, the median interval between initial treatment and diagnosis of the osteosarcomas was 18 years. Those patients who developed postradiation osteosarcoma within 10 years after receiving radiation (latency period < 10 years) fared worse (p <0.0016) than those patients whose latency period was greater than 10 years. The most common histological sub- types varied between studies.16,26 Fibroblastic was the most common histological subtype in our series; however, it was not a prognostic factor for overall survival.
Metastases at presentation carry a poor prognosis. Brady et al8 identified metastases at diagnosis as one of the three adverse prognostic factors. Tabone et al26 reported all patients who presented with metastatic disease died. We only had one patient who presented with lung metastases. The presence of lung metastases at diagnosis did not correlate with decreased survival; however, the number of patients presenting with metastases may have been too small to examine for significance.
All patients underwent surgical resection, limb salvage, resection, or amputation. Negative margins were obtained in all except one patient. Type of surgery (limb sparing versus amputation) or margins obtained did not correlate with local recurrence or overall patient survival. In addition, since the lesions were skeletally based, the entire radiated field was rarely resected. Unlike the experience reported by some authors for radiation sarcomas, there was no correlation between local failure and the type of surgery/amount of radiated tissue resected.
Postradiation osteosarcoma does not appear to have the same prognosis as primary osteosarcomas. Contributing factors may be the advanced age and poor performance status of the patient population. It does not appear the long-term survivors in this study differed from patients who died of disease; however, the difference may be associated with the biology of the tumor. Further investigations on the molecular level are necessary to differentiate long-term survivors and customize treatment regimens. New treatment options, ones not systemically toxic but tumor specific, may be of particular use in the subset of osteosarcoma patients.
We would like to thank Rhonda R. Johnson and Alan Ramirez for their technical assistance in preparing this manuscript.
1. Arlen M, Higinbotham NL, Huvos AG, Marcove RC, Miller T, Shah IC.Radiation-induced sarcoma of bone. Cancer
. 1971;28: 087-1099.
2. Bacci G, Forni C, Ferrari S, Longhi A, Bertoni F, Mercuri M, Donati D, Capanna R, Bernini G, Briccoli A, Setola E, Versari M. Neoadjuvant chemotherapy for osteosarcoma of the extremity: intensification of preoperative treatment does not increase the rate of good histologic response to the primary tumor or improve the final outcome. J Pediatr Hematol Oncol
3. Bacci G, Picci P, Ruggieri P, Mercuri M, Avella M, Capanna R, Brach Del Prever A, Mancini A, Gherlinzoni F, Padovani G, et al. Primary chemotherapy and delayed surgery (neoadjuvant chemo- therapy) for osteosarcoma of the extremities. The Istituto Rizzoli Experience in 127 patients treated preoperatively with intravenous methotrexate (high versus moderate doses) and intraarterial cisplatin. Cancer
4. Benjamin RS. Chemotherapy for osteosarcoma. In: Unni KK, ed. Bone Tumors
. New York: Churchill Livingstone; 1988:149-156.
5. Bielack SS, Kempf-Bielack B, Delling G, Exner GU, Flege S, Helmke K, Kotz R, Salzer-Kuntschik M, Werner M, Winkelmann W, Zoubek A, Jurgens H, Winkler K. Prognostic factors in high- grade osteosarcoma of the extremities or trunk: an analysis of 1,702 patients treated on neoadjuvant cooperative osteosarcoma study group protocols. J Clin Oncol
6. Bielack SS, Kempf-Bielack B, Heise U, Schwenzer D, Winkler K. Combined modality treatment for osteosarcoma occurring as a second malignant disease (Cooperative German-Austrian-Swiss Osteosarcoma Study Group). J Clin Oncol
7. Bielack SS, Tabone MD.Osteosarcomas occurring as second malignant neoplasms. Radiother Oncol
8. Brady MS, Gaynor JJ, Brennan MF. Radiation-associated sarcoma of bone and soft tissue. Arch Surg
9. Buchholz TA, Theriault RL, Niland JC, Hughes ME, Ottesen R, Edge SB, Bookman MA, Weeks JC. The use of radiation as a component of breast conservation therapy in National Comprehensive Cancer Network Centers. J Clin Oncol
10. Cahan WG, Woodard HQ, Higinbotham NL, Stewart FW, Coley BL.Sarcoma arising in irradiated bone: report of eleven cases: 1948. Cancer
11. Clarke M, Collins R, Darby S, Davies C, Elphinstone P, Evans E, Godwin J, Gray R, Hicks C, James S, MacKinnon E, McGale P, McHugh T, Peto R, Taylor C, Wang Y.Effects of radiotherapy and of differences in the extent of surgery for early breast cancer on local recurrence and 15-year survival: an overview of the randomised trials. Lancet
12. Davis AM, Bell RS, Goodwin PJ. Prognostic factors in osteosarcoma: a critical review. J Clin Oncol
13. Frassica FJ, Frassica DA, Wold LE, Beabout JW, Sim FH. Postradiation sarcoma of bone. Orthopedics
. 1993;16:105-106, 109.
14. Hamre MR, Severson RK, Chuba P, Lucas DR, Thomas RL, Mott MP.Osteosarcoma as a second malignant neoplasm. Radiother Oncol
15. Healey JH, Buss D.Radiation and pagetic osteogenic sarcomas. Clin Orthop Relat Res
16. Huvos AG, Woodard HQ, Cahan WG, Higinbotham NL, Stewart FW, Butler A, Bretsky SS.Postradiation osteogenic sarcoma of bone and soft tissues: a clinicopathologic study of 66 patients. Cancer
17. Meyers PA, Heller G, Healey J, Huvos A, Lane J, Marcove R, Applewhite A, Vlamis V, Rosen G. Chemotherapy for nonmeta- static osteogenic sarcoma: the Memorial Sloan-Kettering experience. J Clin Oncol
18. Mindell ER, Shah NK, Webster JH.Postradiation sarcoma of bone and soft tissues. Orthop Clin North Am
19. Mole RH.Late Effects of Radiation Carcinogenesis. Br. Med. Bull
20. Murray EM, Werner D, Greeff EA, Taylor DA.Postradiation sarcomas: 20 cases and a literature review. Int J Radiat Oncol Biol Phys
21. Pratt CB, Meyer WH, Rao BN, Pappo AS, Fleming ID, Luo X, Cain A, Kaste SC, Shearer PD, Jenkins JJ,3rd. Comparison of primary osteosarcoma of flat bones with secondary osteosarcoma of any site. Cancer
22. Robinson E, Neugut AI, Wylie P. Clinical aspects of post- irradiation sarcomas. J Natl Cancer Inst
23. Seeler R.Book Review: Pediatric Oncology: A Treatise for the Clinician. JAMA
24. Sim FH, Cupps RE, Dahlin DC, Ivins JC. Postradiation sarcoma of bone. J Bone Joint Surg Am
25. Simon MA, Aschliman MA, Thomas N, Mankin HJ. Limb-salvage treatment versus amputation for osteosarcoma of the distal end of the femur. J Bone Joint Surg Am
26. Tabone MD, Terrier P, Pacquement H, Brunat-Mentigny M, Schmitt C, Babin-Boilletot A, Mahmoud HH, Kalifa C. Outcome of radiation-related osteosarcoma after treatment of childhood and adolescent cancer: a study of 23 cases. J Clin Oncol
. 1999;17: 2789-2795.
27. Tountas AA, Fornasier VL, Harwood AR, Leung PM.Postirradiation sarcoma of bone: a perspective. Cancer
28. Wilkins RM, Cullen JW, Camozzi AB, Jamroz BA, Odom L.Improved survival in primary nonmetastatic pediatric osteosarcoma of the extremity. Clin Orthop Relat Res
29. Wilkins RM, Cullen JW, Odom L, Jamroz BA, Cullen PM, Fink K, Peck SD, Stevens SL, Kelly CM, Camozzi AB. Superior survival in treatment of primary nonmetastatic pediatric osteosarcoma of the extremity. Ann Surg Oncol
30. Winkler K, Beron G, Delling G, Heise U, Kabisch H, Purfurst C, Berger J, Ritter J, Jurgens H, Gerein V, et al. Neoadjuvant chemo- therapy of osteosarcoma: results of a randomized cooperative trial (COSS-82) with salvage chemotherapy based on histological tumor response. J Clin Oncol