ESTRO 2021 Abstract Book

S805

ESTRO 2021

Human papilloma virus (HPV) status, clinical lymph node (N) involvement, and early response to treatment are of key importance for survival prediction in patients in locally advanced oropharyngeal cancer treated with definitive chemoradiotherapy (CRT). We propose the use of a dynamic nomogram to translate estimated survival probability into an informative tool to help both clinicians and patients in decision-making. Materials and Methods Clinical information from 273 patients with histologically confirmed squamous cell carcinoma of the oropharynx treated with definitive chemoradiotherapy from March 2010 to December 2016 were retrospectively reviewed. A dynamic nomogram was developed, using DynNom package in open-source R software. Results Based on the important variables – HPV status, N classification and early response to CRT – previously identified and reported with machine learning-based methodology, it was possible to construct a dynamic nomogram to graphically predict survival. The Figure shows an example. The three variables are: HPV (0 = negative; 1 = positive), early responders (0 = no complete response within 3 months the end of CRT; 1 = complete response within 3 months the end of CRT) and N classification according to 8 th edition TNM. Inputting patient characteristics – for instance HPV 0, Early_responders 1 and N_8 th 2c – allows calculation of the estimated survival probability (blue curve). The black curve refers to a patient HPV 0, early responder and N3b.

Conclusion In the oropharyngeal cancer scenario, this dynamic nomogram has the potential to make survival clinical data easily accessible and case-tailored. External prospective independent validation of the model is planned.

PO-0972 Pre-treatment haemoglobin as a surrogate marker for survival in head & neck squamous cell carcinoma S. Tandon 1 , M. Gairola 2 , P. Ahlawat 2 , S. Purohit 3 , P. Umesh 2 , A. Krishnan 2 1 Rajiv Gandhi Cancer Institute and Research Centre, Radiaton Oncology, Delhi, India; 2 Rajiv Gandhi Cancer Institute and Research Centre, Radiation Oncology, Delhi, India; 3 Rajiv Gandhi Cancer Institute and Research Centre, Radiation Oncology, Dellhi, India Purpose or Objective Hypoxia has an adverse effect on the radiosensitivity of cells. It is believed that oxygen enhances the efficacy of radiotherapy by making DNA damage caused by radiation-induced free radicals less repairable. To exert a radiosensitizing effect, oxygen must be present during radiation exposure at the tumor site. Haemoglobin which is our innate oxygen serves that purpose and in turn may help us in prognosticating the disease. A retrospective review to study the effect of baseline hemoglobin of SCC-HNC diagnosed patients as a prognostic factor/predictive factor of patients being treated with definitive radiation. Primary endpoint is overall survival (OS) and locoregional recurrence free survival (LRFS) at 1 year. Secondary endpoint was LRFS and OS at 1.5 years. Materials and Methods Hundred patients of proven SCC-HNC who were treated with definitive RT were selected between the periods of January’18 to May’19. All patients had their response rates assessed at 12 weeks of completion of radiotherapy. OS and LRFS was calculated using KM curves. Results The median follow-up was 19 months (1 – 39 months) for the entire cohort. A logistic regression was performed to ascertain the effect of Hb on the likelihood that patients were surviving. The results showed, for each unit reduction in the Hb level, the odds of surviving increases by a factor of 1.34 (Odd Ratio 0.742; CI = 0.593 - 0.927; p = 0.009). A receiver operating characteristics (ROC) analysis was performed to determine an