Trends and variability in survival among patients with osteosarcoma: a 7-year update.

This report is an update of a 1978 article on osteosarcoma in Mayo Clinic patients. It includes additional follow-up on previously reported cases and incorporates new cases treated since the time of that original study. From 1963 through 1981, 336 patients with classic, previously untreated osteosarcoma received their first definitive treatment at our institution. Survival of these patients was studied in detail. The most significant result was that survival in the 1960s was much worse than that in the 1970s. The first evidence of improvement in survival was noted in 1969; subsequently, further improvement occurred but was not consistent. This finding prevailed with respect to duration of survival to death, survival to detection of metastasis, and survival from occurrence of metastasis to death. On the basis of detailed regression analysis, several variables had independent prognostic value. From these findings, a prognostic score was developed, which was based on the number of the following unfavorable characteristics: age younger than 10 years, male sex, tumor diameter more than 15 cm, cell type osteoblastic or chondroblastic, duration of symptoms 2 months or less, and involvement of the femur or humerus. Patients with five or six of these unfavorable characteristics had a very poor survival; in contrast, patients with only one or two characteristics had a good outcome. Even when these scores were fairly constant, however, the calendar period had a strong influence on survival. Likewise, when treatment was considered and adjustments by score were made, no significant differences could be found between those patients treated by amputation only and those treated by amputation supplemented with chemotherapy or radiotherapy.