The impact of chemo- and radiotherapy treatments on selfish de novo FGFR2 mutations in sperm of cancer survivors.

study questionWhat effect does cancer treatment have on levels of spontaneous selfish fibroblast growth factor receptor 2 (FGFR2) point mutations in human sperm? SUMMARY ANSWER: Chemotherapy and radiotherapy do not increase levels of spontaneous FGFR2 mutations in sperm but, unexpectedly, highly-sterilizing treatments dramatically reduce the levels of the disease-associated c.755C > G (Apert syndrome) mutation in sperm. WHAT IS KNOWN ALREADY: Cancer treatments lead to short-term increases in gross DNA damage (chromosomal abnormalities and DNA fragmentation) but the long-term effects, particularly at the single nucleotide resolution level, are poorly understood. We have exploited an ultra-sensitive assay to directly quantify point mutation levels at the FGFR2 locus. STUDY DESIGN, SIZE, DURATION: 'Selfish' mutations are disease-associated mutations that occur spontaneously in the sperm of most men and their levels typically increase with age. Levels of mutations at c.752-755 of FGFR2 (including c.755C > G and c.755C > T associated with Apert and Crouzon syndromes, respectively) in semen post-cancer treatment from 18 men were compared to levels in pre-treatment samples from the same individuals (n = 4) or levels in previously screened population controls (n = 99). PARTICIPANTS/MATERIALS, SETTING,

methodsCancer patients were stratified into four different groups based on the treatments they received and the length of time for spermatogenesis recovery. DNA extracted from semen samples was analysed using a previously established highly sensitive assay to identify mutations at positions c.752-755 of FGFR2. Five to ten micrograms of semen genomic DNA was spiked with internal controls for quantification purposes, digested with MboI restriction enzyme and gel extracted. Following PCR amplification, further MboI digestion and a nested PCR with barcoding primers, samples were sequenced on Illumina MiSeq. Mutation levels were determined relative to the spiked internal control; in individuals heterozygous for a nearby common single nucleotide polymorphism (SNP), mutations were phased to their respective alleles. MAIN RESULTS AND THE ROLE OF CHANCE: Patients treated with moderately-sterilizing alkylating regimens and who recovered spermatogenesis within <3 years after therapy (Group 3, n = 4) or non - alkylating chemotherapy and/or low gonadal radiation doses (Group 1, n = 4) had mutation levels similar to untreated controls. However, patients who had highly-sterilizing alkylating treatments (i.e. >5 years to spermatogenesis recovery) (Group 2, n = 7) or pelvic radiotherapy (Group 4, n = 3) exhibited c.755C > G mutation levels at or below background. Two patients (A and B) treated with highly-sterilizing alkylating agents demonstrated a clear reduction from pre-treatment levels; however pre-treatment samples were not available for the other patients with low mutation levels. Therefore, although based on their age we would expect detectable levels of mutations, we cannot exclude the possibility that these patients also had low mutation levels pre-treatment. In three patients with low c.755C > G levels at the first timepoint post-treatment, we observed increasing mutation levels over time. For two such patients we could phase the mutation to a nearby polymorphism (SNP) and determine that the mutation counts likely originated from a single or a small number of mutational events. LIMITATIONS, REASONS FOR CAUTION: This study was limited to 18 patients with different treatment regimens; for nine of the 18 patients, samples from only one timepoint were available. Only 12 different de novo substitutions at the FGFR2 c.752-755 locus were assessed, two of which are known to be disease associated. WIDER IMPLICATIONS OF THE

findingsOur data add to the body of evidence from epidemiological studies and experimental data in humans suggesting that male germline stem cells are resilient to the accumulation of spontaneous mutations. Collectively, these data should provide physicians and health-care professionals with reassuring experimental-based evidence for counselling of male cancer patients contemplating their reproductive options several years after treatment. STUDY FUNDING/COMPETING INTEREST(S): This work was primarily supported by grants from the Wellcome (grant 091182 to AG and AOMW; grant 102 731 to AOMW), the University of Oxford Medical Sciences Division Internal Fund (grant 0005128 to GJM and AG), the National Institute for Health Research (NIHR) Oxford Biomedical Research Centre Programme (to AG) and the US National Institutes of Health (to MLM). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. None of the authors has any conflicts of interest to declare. TRIAL REGISTRATION NUMBER: NA.