Faculty Bibliography

Members of the bone morphogenetic protein (BMP) family and their receptors (BMPRs and activin receptors-ActRs) promote the development of bones with a fine regulation of their expression. Mutations in BMPs or BMPRs cause several diseases, as shown in knockout mice, such as skeletal defects, familial primary pulmonary hypertension and neoplasias. Osteosarcoma is the most frequent primary malignant tumor of bone. Due to their importance in bone development, BMPs, BMPRs and ActRs could also play a role in osteosarcoma growth and development. Previous data have shown that the overexpression of the BMPR-II was related to poor prognosis in malignant and metastatic bone tumors. We evaluate by reverse transcription-linked polymerase chain reaction analysis (RT-PCR) the expression pattern of BMPs, BMPRs and ActRs in five different human osteosarcoma cell lines (MG63, G292, HOS, SaOS and U2). Moreover, we performed the mutational screening of the complete BMPR-II mRNA by automated sequencing of the correspondent cDNA to evaluate the presence of point mutations in osteosarcoma cell lines. All the osteosarcoma cell lines studied simultaneously expressed the BMPs, BMPRs and ActRs investigated. No mutations were detected in the BMPR-II cDNA. Our results suggest the presence of a mechanism involving the simultaneous activation of the BMPs and their receptors in osteosarcoma cell lines.

Between January 1995 and December 1999, 11 patients with synchronous multifocal osteosarcoma (SMO) received neoadjuvant treatment with high-dose methotrexate, cisplatinum, Adriamycin, and ifosfamide. After primary chemotherapy in 4 patients who had only two bone localizations, it was possible to treat all tumor foci locally. The remaining patients, with more than three bones involved, were treated surgically only in 3 cases at the primary site, while secondary lesions did not receive any treatment. The final results of our study were disappointing. All patients died of the tumor 6 to 24 months after the beginning of treatment (mean 11.9 months). Nevertheless, the survival time of the 4 patients with locally treated lesions was significantly longer than the one of 7 patients in whom the secondary lesions were not locally treated (18.2 vs 9.1 months; P<0.008). It should be noted that those patients simultaneously operated on two sites, the response to chemotherapy of "primary" and "secondary" lesions was always similar. This homogeneity supports the thesis that in synchronous multifocal osteosarcoma the tumors are not multicentric in origin but represent bone-to-bone metastases from a monocentric tumor.

Right-sided colon cancer therapy is based upon right hemicolectomy, which is a well established surgical technique. Considering the possibility of biological diversity in cancer arising from the caecum, ascending colon and transverse colon, this retrospective study analysed age, sex, grading, T, N, M and mortality to evaluate the biological characteristics of the cancers and their level of local and general diffusion with a view to establishing any correlation between these characteristics and the advisability of employing different surgical procedures for the different sites. Cancer arising from the sites considered does not seem to present different biological aspects and therefore a different surgical procedure is not authorized. Our data show that a significant majority of cancers are diagnosed at an advanced stage as compared to earlier stages (P < 0.05) and that early detection of these cancers is the best way of improving the prognosis.

Between April 1990 and December 1994, we treated 24 patients with telangiectatic osteosarcoma (TO) of the extremities with neoadjuvant chemotherapy using 2 protocols. Surgery consisted of limb salvage in 21 patients and amputation or rotation plasty in 3. The histologic response to chemotherapy was good (90% or more tumor necrosis) in 23 patients, of whom 12 had total necrosis. With a mean follow-up of 74 (60-96) months, 20 patients remained continuously free of disease and 4 relapsed with lung metastases. There were no local recurrences. Comparing these results to the ones achieved in 269 contemporary patients with conventional osteosarcoma of the extremities using the same protocols for chemotherapy, we found a significantly better histologic response to chemotherapy (96% vs 68% of good histologic response; p = 0.004) and disease-free survival (83% vs 55%; p = 0.01) in the TO group. We conclude that TO, once considered a lethal tumor, seems to be even more sensitive to chemotherapy than conventional osteosarcoma, and that most of these patients may be cured without amputation.

The results of the Rizzoli IOR/OS-3b neoadjuvant protocol for the treatment of osteosarcoma of the extremity are reported. Preoperative chemotherapy consisted of two cycles of high-dose methotrexate (HDMTX i.v.), followed by a combination of cisplatin (CDP i.a.)/ doxorubicin (ADM i.v.). Postoperatively all patients, regardless of the histologic response, received 3 more cycles of MTX, CDP/ADM alternated with 3 cycles of ifosfamide. In the study performed between January and December 1992 43 patients were enrolled and limb salvage was performed in 39 of them (91%). The histologic response to chemotherapy was good (90% or more tumor necrosis) in 24 patients (56%) and poor (less than 90% tumor necrosis) in 19 (44%). With a minimum follow-up of 7 years, 23 pts (53%) remained continuously free of disease, 19 relapsed and one died due to unrelated cause. In spite of the high number of limb salvages performed, only 2 local recurrences were registered. The 7-year event-free survival and overall survival were, respectively, 53% and 68%. The hematopoietic and extrahematopoietic toxicity experienced by the patients during the entire treatment was relatively mild. These long-term results confirm that, with neoadjuvant chemotherapy, it is possible to cure more than 60% of patients with osteosarcoma of the extremities, avoiding amputation in most of them. These results, however, are no better than those achieved in our previous study IOR/OS-3a, in which only poor responder patients received ifosfamide during the postoperative treatment.

570 patients with osteosarcoma of the extremities were treated with five different protocols of neoadjuvant chemotherapy at Rizzoli Institute between 1983 and 1995. Surgery consisted of limb salvage in 83% rotation plasty in 5% and amputation in 12%. The 5-year event-free survival (EFS) was 60% which varied according to the protocol followed, ranging from 47.6% to 66.4%. 234 patients relapsed. The pattern of relapse was analysed. The mean relapse time was 23.8 months (range: 2-96). The first site of systemic relapse was the lung in 88% (32% of these had less than three pulmonary metastases and 68% three or more), bone in 9%, lung and bone in 2% and other sites in 3%. The relapse time and the number of pulmonary metastases were strictly correlated with the efficacy of the protocol of chemotherapy used. Patients treated with the three protocols that gave a 5-year EFS of more than 60% relapsed later and had fewer pulmonary lesions than patients treated with the two protocols that gave a 5-year EFS of 47.6% and 52.5%. The rate of local recurrence was relatively low (6%). This was not correlated with the protocol or the type of surgery used: limb salvage (6.4%), rotation plasty or amputation (4.1%). However, the rate of local recurrence was very high (21.9%) in the few patients (7%) that had less than wide surgical margins. We conclude that for patients with osteosarcoma of the extremities treated with neoadjuvant chemotherapy: (a) the pattern of systemic relapse changes according to the efficacy of the protocol of chemotherapy used. This should be always considered when evaluating the preliminary results of new studies as well as in defining the time of follow-up; (b) limb salvage procedures are safe and do not jeopardise the outcome of the patient, provided that wide surgical margins are achieved.