ABSTRACT
PURPOSE: Most radiobiological models for prediction of tumor control probability (TCP) do not account for the fact that many events could remain unobserved because of censoring. We therefore evaluated a set of TCP models that take into account this censoring. METHODS AND MATERIALS: We applied 2 fundamental Bayesian cure rate models to a sample of 770 pulmonary metastasis treated with stereotactic body radiation therapy at German, Austrian, and Swiss institutions: (1) the model developed by Chen, Ibrahim and Sinha (the CIS99 model); and (2) a mixture model similar to the classic model of Berkson and Gage (the BG model). In the CIS99 model the number of clonogens surviving the radiation treatment follows a Poisson distribution, whereas in the BG model only 1 dominant recurrence-competent tissue mass may remain. The dose delivered to the isocenter, tumor size and location, sex, age, and pretreatment chemotherapy were used as covariates for regression. RESULTS: Mean follow-up time was 15.5 months (range: 0.1-125). Tumor recurrence occurred in 11.6% of the metastases. Delivered dose, female sex, peripheral tumor location and having received no chemotherapy before RT were associated with higher TCP in all models. Parameter estimates of the CIS99 were consistent with the classical Cox proportional hazards model. The dose required to achieve 90% tumor control after 15.5 months was 146 (range: 114-188) Gy10 in the CIS99 and 133 (range: 101-164) Gy10 in the BG model; however, the BG model predicted lower tumor control at long (â³20 months) follow-up times and gave a suboptimal fit to the data compared to the CIS99 model. CONCLUSIONS: Biologically motivated cure rate models allow adding the time component into TCP modeling without being restricted to the follow-up period which is the case for the Cox model. In practice, application of such models to the clinical setting could allow for adaption of treatment doses depending on whether local control should be achieved in the short or longer term.
Subject(s)
Lung Neoplasms/secondary , Lung Neoplasms/therapy , Models, Theoretical , Radiosurgery/methods , Adolescent , Adult , Aged , Aged, 80 and over , Bayes Theorem , Child , Databases, Factual , Female , Follow-Up Studies , Humans , Lung Neoplasms/mortality , Male , Middle Aged , Models, Biological , Neoplasm Recurrence, Local , Poisson Distribution , Probability , Radiotherapy Dosage , Retrospective Studies , Time Factors , Young AdultABSTRACT
BACKGROUND AND PURPOSE: To evaluate whether local tumor control probability (TCP) in stereotactic body radiotherapy (SBRT) varies between lung metastases of different primary cancer sites and between primary non-small cell lung cancer (NSCLC) and secondary lung tumors. MATERIALS AND METHODS: A retrospective multi-institutional (n=22) database of 399 patients with stage I NSCLC and 397 patients with 525 lung metastases was analyzed. Irradiation doses were converted to biologically effective doses (BED). Logistic regression was used for local tumor control probability (TCP) modeling and the second-order bias corrected Akaike Information Criterion was used for model comparison. RESULTS: After median follow-up of 19 months and 16 months (n.s.), local tumor control was observed in 87.7% and 86.7% of the primary and secondary lung tumors (n.s.), respectively. A strong dose-response relationship was observed in the primary NSCLC and metastatic cohort but dose-response relationships were not significantly different: the TCD90 (dose to achieve 90% TCP; BED of maximum planning target volume dose) estimates were 176 Gy (151-223) and 160 Gy (123-237) (n.s.), respectively. The dose-response relationship was not influenced by the primary cancer site within the metastatic cohort. CONCLUSIONS: Dose-response relationships for local tumor control in SBRT were not different between lung metastases of various primary cancer sites and between primary NSCLC and lung metastases.
