Background Men with castrate resistant prostate cancer (CRPC) suffer from symptoms related to both their disease and its treatment with marked variation between individuals with respect to symptom expression despite similar tumour burdens. This may be related to genetic variation.

Objective This study aimed to determine whether genetic variation in cytokine expression can predict severity of symptoms in men with CRPC and investigate whether symptom severity was related to tumour burden.

Methods A prospective, longitudinal consecutive patient cohort study across two Queensland sites. Patient characteristics including tumour burden and current treatment were collected at baseline. Symptom severity was assessed using the Edmonton Symptom Assessment Scale (ESAS-R) 3–4 weekly for up to 6 assessments, with blood taken for genetic analysis once during the study. Cytokine gene variants of each participant were assessed using a panel of SNPs most reported to be associated with symptom variation in the literature. Analysis was performed using R software and the package SNPassoc (R Core team, 2021; Gonzalez JR. et al, 2007).

Results Twenty-eight of 67 (42%) of participants had a low, and 39/67 (58%) a high tumour burden. Tumour burden remained constant throughout the study period for all participants. Twenty-six of 67 (39%) participants were classified as having low ESAS-R severity and 67 (61%) as high ESAS-R severity. Symptom severity was not related to tumour burden or patient characteristics. One hundred and forty-four SNPs from 63 participants were analysed, of which results were available for 142 SNPs. Fifteen SNPs were significantly associated with symptom severity. In multivariable analysis, SNP rs2069772 in IL2 (p = 0.001), rs1554606 in IL6 (p=0.003), rs2227306 in IL8 (p=0.008) and rs2069718 in IFNG (p=0.001) were highly predictive of symptom severity.

Discussion Although multiple factors can influence symptom severity, genetic variation may well play a part. The early identification of men likely to develop severe symptoms during the course of their prostate cancer could theoretically enable symptoms to be managed more aggressively from an early stage. Our findings also add to the expanding data bases that document the associations of genetic polymorphism with both disease and symptom expression.