Cellular and molecular mechanisms of response to, and recurrence following, breast radiotherapy

Abstract

This thesis explores response to, and recurrence following, radiotherapy in primary breast cancer within three clinical trials investigating aspects of radiation management: the IMPORT trials, IMPORT LOW and IMPORT HIGH, and the NeoRT trial. Clinical outcomes from the NeoRT trial are described as well as the results of translational analyses of patient samples from all 3 trials, aiming to interrogate genomic aspects of radiotherapy response and of recurrence following breast conserving surgery and radiotherapy. Ipsilateral breast tumour recurrences following breast conserving treatment have long been noted to predominantly occur in or close to the tumour bed from which the index cancer was removed with a smaller number arising in other quadrants of the breast. A widely held hypothesis is that breast cancer local recurrence after partial breast radiotherapy usually represents ‘true recurrence’ – i.e. derived from residual cells of the index cancer - if arising within the quadrant of the original tumour, and an independently arising new primary cancer if arising beyond this index quadrant or in the contralateral breast. The IMPORT trials are both large UK randomised controlled trials investigating the paradigm of adapting radiotherapy dose/fractionation and volume to the spatially varying risk of recurrence across the breast. Partial breast radiotherapy is investigated in a cohort at low risk of local recurrence in IMPORT LOW. Simultaneous integrated boost to two dose escalation levels is compared to a control of sequential radiotherapy boost in a cohort at higher risk of local recurrence in the sister trial, IMPORT HIGH. This work analyses the genomic, spatial and clinicopathological relationships between paired index and subsequent tumours within the two trials, aiming to answer the following questions:

Can the Breakclone methodology, which integrates analysis of shared and private mutations and copy number abnormality breakpoints in tumour pairs as well as cohort and population frequencies of these abnormalities, classify tumours diagnosed as IBTR in the IMPORT trials as either true recurrence or new primary? What are the spatial and temporal patterns of true recurrence and new primary across the two IMPORT cohorts? How do clinical outcomes differ between those classified as true recurrence and new primary via this method? The most notable findings were that ipsilateral breast tumour recurrences in IMPORT LOW were significantly less likely to be classed as clonally related than those in IMPORT HIGH; 13/21 (62%) were judged to be new primary cancers. Additionally a subset of contralateral cancers in the women with high-risk cancers were genomically highly similar to the index tumours, suggesting that these are likely metastases. Contrary to expectations there was no clear spatial pattern to the location of the subsequent tumour relative to the index tumour bed. Higher deletion:insertion and indel:substitution ratios were detected in ipsilateral subsequent tumours in the analysis, in keeping with previous reporting in radiation-associated second malignancies. Amplification of 8q24.21, on which c-myc is located was more frequent in those cases that developed a clonal recurrence. v Patients undergoing breast conserving surgery, as opposed to mastectomy, have better body image and quality of life scores; eligibility for the former depends in part upon the size of the tumour relative to the size of the breast, and interventions that can reduce tumour size may increase the likelihood of breast conservation. The NeoRT trial is a feasibility study of neoadjuvant breast radiotherapy and endocrine therapy in oestrogen receptor positive breast cancer, for whom this might facilitate more conservative surgery. The primary endpoint of the trial was feasibility, defined by the proportion of patients successfully completing neoadjuvant radiotherapy and endocrine treatment followed by breast surgery; clinical secondary endpoints including mastectomy rate, peri/post-operafve complicafons, pathological response to treatment and late normal fssue toxicity. Translational opportunities were built into the trial protocol: sequential tumour research biopsies and research blood samples were taken at four timepoints. These were used to explore circulating tumour DNA (ctDNA) kinetics as well as changes in the genomic landscapes over the course of treatment. This work therefore aims to answer the following questions: Is the schedule of preoperative radiotherapy followed by 20 weeks endocrine therapy feasible and safe in this context, and is there any indication of a reduction in mastectomy rate? Is it possible to detect a biological signal of response to radiotherapy via: a. Genomic features? b. Circulating tumour DNA kinetics or composition? Although the number of patients recruited to the trial was less than planned, statistical analysis suggested that the primary endpoint of feasibility would have been met if recruitment was completed. Tumour informed circulating tumour DNA analysis was positive in several cases/timepoints, with the highest tumour fractions detected in a case in which the patient was found to have metastatic disease, but other than this without any clear associations with clinicopathological parameters. Marked heterogeneity in tumour mutational profiles was seen between timepoints in the majority of cases; this could potentially be attributed to spatial differences in biopsy acquisition. The number of small deletions detected was however significantly higher at later timepoints compared to baseline, again in keeping with a previous study of radiation-associated malignancies. The IMPORT trials recurrence study findings that many subsequent cancers following low risk breast cancer are in fact new primaries, and that a subset of contralateral tumours following high risk breast cancer seem to be metastases, will have clear implications implications for management of these cases if validated. The NeoRT clinical trial findings add to evidence for the feasibility of neoadjuvant breast radiotherapy.