Osteosarcoma is the most common malignant bone tumour in children and young people. Approximately 40% patients respond poorly to highly toxic preoperative MAP (methotrexate adriamycin, cisplatin) chemotherapy with consequent inferior survival. The role of genetic polymorphisms in drug response and toxicity is reported in acute leukaemia and some solid tumours. Recent evidence in osteosarcoma suggests increased chemotherapy toxicity is associated with improved survival. The aim of this pilot study is to investigate the influence of drug target and metabolising gene polymorphisms on tumour response and chemotherapy toxicity in osteosarcoma.

Patients who have completed MAP chemotherapy are eligible. Chemotherapy toxicity (CTCAE grade) is collected from patient records. Tumour response is graded as good (> 90% necrosis) or poor (< 90% necrosis) in resection specimen. Peripheral blood DNA is typed for genome-wide single nucleotide polymorphisms (SNP) using the Illumina 610 Quad array and analysed using Bead Studio software. Standard PCR techniques are used to genotype the Thymidylate synthase (TS) gene (folate pathway) for the presence of 2 or 3 copies of a 28 base pair repeat (2R/3R) and a G/C SNP in the 3R allele.

52 patients have entered to date: 33 good responders, 12 poor and 7 unevaluable for response. Median age 18 years (range 10–51), males:females 1.3:1. Median follow up is 39 months (range 2–76) with 11 patients relapsing. 23 patients have TS genotype 2R/2R, nineteen 2R/3R, six 3R/3R, three 2R/4R and one 2R/7R. Neither TS repeat or G/C SNP genotype correlated with histological response or degree of methotrexate stomatitis. Interestingly, presence of the 2R allele was significantly related to relapse (p=0.01) but may reflect small patient numbers. Recurrent methotrexate stomatitis (> 2 episodes of CTCAE grade 2) was weakly correlated with no relapse (p=0.07). Analysis of SNP array data with emphasis on MAP pathway polymorphisms will be presented when complete.

Medical records of children < /=5 years, treated by the London sarcoma service for malignant primary bone tumours (average new cases osteosarcoma (OS)/Ewings sarcoma (ES), all ages: 125/year) between 1999 and 2009, were reviewed.

Results: 5 OS and 6 ES. Mean age – 4.2 years (range 2.1–5.8), 8/11 males. OS primary sites: distal femur (2), proximal femur (1) and proximal humerus (2); localised tumours only. Primary sites in the ES cohort included 1 distal femur, 2 chest wall (1 – spinal extension), 1 buttock (spinal extension), 1 temporal bone and 1 ulna; 1 had bone/bone marrow involvement, 1 had chest metastases. 4/5 OS (Euramos, MRC B007) and 5/6 ES (Euro-Ewings 99) were entered into phase III clinical trials. Delayed surgery for OS occurred at mean 12.1 weeks (range 11–13) – 4 limb salvage prostheses with 2/4 non-invasive growers, 1 forequarter amputation. All had a good (> 90% necrosis) histologic response to neoadjuvant therapy. Delayed surgery for ES occurred at mean 21.7 weeks (range 12.8 – 35), 1 limb salvage with prosthesis (non-invasive grower), 1 biological reconstruction; remainder had tumour resections. Histologic response: 50% good. In the OS cohort, 1 child died a toxic death; 1 developed pulmonary metastases and died 2 years from diagnosis; 1 has a metastatic recurrence in the opposite humerus 2 years from diagnosis and starts 2nd line therapy; 1 had local recurrence 1 year from diagnosis but alive at 7.4 years;1 alive/disease free at 2.5 years. In the ES cohort 5/6 are alive disease free -1, 4.1, 5.2, 6.9 and 7 years from diagnosis; 2 needed 2nd line therapy for recurrent distant disease 4.5 and 5.8 years off therapy, 1 of whom has just recurred again (6.1 years from diagnosis).

Conclusion- improving early survival rates in the very young with OS remains a significant challenge. Quality of survival requires further age-appropriate study.