Radiation Induced Lymphocyte Apoptosis Assay and Predictive Breast Fibrosis Nomogram: First Implementation in Daily Practice in France

Abstract

The radiation-induced CD8-lymphocyte apoptosis assay (RILA) adjusted by the NovaGray Breast® nomogram has been shown to significantly predict breast fibrosis risk. Recent prospective trials confirmed the predictive role of RILA in urinary, sexual or anorectal toxicity risk after prostate radiotherapy. This study aims at evaluating the daily use of this assay and its impact on current practice. Between 2007 and 2016, 304 patients benefited from the RILA assay outside of clinical trial setting. We analyzed the reasons for assay prescription, late toxicity (assessed by CTCAEv3.0 grading scale) and clinical impact of the result. For non-breast demands, RILA cutoffs for high and low values were taken from the results of the REQUITE trial and Ozsahin et al. (Clin Cancer Res, 2005) For breast tumors, all RILA results were adjusted by the nomogram developed by NovaGray based on the recent results of the multicenter prospective trial (Azria et al., Ebiomedecine 2015). 149 patients with sufficient available data were included in the analysis. 51 patients were assessed because of late toxicities after radiotherapy (RT). Of these, 36 (70.6%) had low and 15 (29.4%) intermediate/high RILA values. For breast cancer, 27 of 29 patients with toxicities (93.1%) fitted with the predictive nomogram. 13 patients were assessed before re-irradiation, and 2 saw their RT treatment cancelled because of low RILA values (one had surgery and one refused treatment). 85 patients were assessed before initial RT treatment. Of these, 36 (42.4%) were assessed because of medical history (i.e.: systemic sclerosis, lupus or polymorphous light eruption), 33 (38.9%) because the patient asked for the test, 7 (8.2%) to aid clinical decision (i.e.: proximity of organs at risk or unclear RT indication), 4 (4.7%) because of history of surgery or trauma in the treated area, 4 (4.7%) before inclusion in clinical trials which they ended excluded from, and one (1.2%) because of severe toxicity after chemotherapy. 23 of the 85 patients (27%) had treatment changes because of the test results. Seven patients with low RILA (30.4%) saw their RT treatment replaced by other treatment modalities (mastectomy, radical prostatectomy, focal treatment, hormone therapy or brachytherapy). Treatment was canceled altogether because of low RILA in 5 cases (21.8%). Eight patients (34.8%) with low RILA had changes in the RT regimen (2 total dose reduction, 3 fractionation change, 3 RT modality change towards IMRT). Three patients (13%) had the RT indication maintained despite the test results. This is the first daily clinical implementation of the RILA assay and breast nomogram risk. Interestingly, half of the patients were assessed before treatment, with a high rate of treatment change based on the results, showing the importance of predictive markers before treatment decision.