Differences in Fecal Microbiota in Long-Term Adolscent/Young Adult Hodgkin Lymphoma Survivors and Their Unaffected Twins

Background: Hodgkin lymphoma is one of the most common malignancies among adolescents and young adults (AYA) in economically developed settings. Although the survival rate is very high (85%), severe late effects are common. Immune dysregulation is clinically and pathologically associated with HL at diagnosis and alterations in inflammatory cytokine levels are evident years after cure. Gut bacteria play a role in modulating immune responses and inflammation. In a pilot study of 13 AYA HL-discordant twin pairs, we saw significantly lower fecal microbial diversity in the HL survivor compared to their unaffected twin. We conducted a follow-up study in a larger set of twins to determine whether there are differences in the relative abundance of specific fecal microbiota in long-term AYAHL survivors compared to their unaffected twins. Twin pairs are an ideal study population because they are partially matched on genetic and early life factors, both of which influence the composition of the gut microbiome.

Methods: The subjects are members of identical twin pairs enrolled in the USC International Twin Study, in whom one twin was diagnosed with histologically verified AYA HL diagnosed at 15-50 years old, at least 10 years ago (mean interval since diagnosis = 22 years ago), and the other twin has remained unaffected. 24 AYA HL survivors and their unaffected twins provided stool samples and detailed answers to an online questionnaire on timing of life milestones, quality of life and medical conditions. A single unaffected co-twin also participated; her twin (the AYA HL survivor) was ineligible because of recent antibiotic use. Fresh stool samples were collected, frozen at -80⁰, shipped to our laboratory and maintained at that temperature until use. DNA extraction was performed using bead beating in conjunction with the MO BIO Powersoil kit. Sequencing of the 253 base pair V4 region of 16S ribosomal DNA was performed using an Illumina HiSeq 2500. Raw sequence data was processed in QIIME and 97% OTUs picked using the Silva database. Taxa present in less than 10% of samples were excluded. Alpha diversity metrics including Chao1 index, Shannon index, and Faith's phylogenetic diversity were calculated for each sample and differences tested with a paired t-test. Sequences were rarefied with 1,000 samplings of the lowest number of reads to standardize the sample and an average was taken to calculate alpha diversity. Differential genera between AYA HL survivors and their unaffected twins were determined using DESeq2, based upon negative binomial models with twin pair as a covariate. P-values were adjusted for multiple comparisons.

Results: Sequencing yielded an average of 219,373 reads, with a minimum of 64,742 and a maximum of 391,120. There were no significant differences between AYA HL survivors and their unaffected twins in the various alpha diversity metrics. However for each measure, the unaffected twin had a higher diversity score compared with their HL case twin survivor. There were no differences in beta-diversity. Significant differences in relative abundance of the following genera were observed between AYA HL survivors and their unaffected twins: Streptococcus (p=9.7x 10^-5), Sellimonas (p= 0.005), Candidatus Stoquefichus (p=0.020), unclassified S24-7 (p=0.035), Faeclitalea (p=0.035), Veillonella (p=0.040), Faecalibacterium (p=0.026) and [Eubacterium] oxidoreducens group (p=0.028). The last two taxa were more abundant in survivors compared with unaffected twin and the remaining taxa were more abundant in unaffected twin controls compared with survivors.

Conclusions: Even after decades, AYA HL survivors harbor fecal microbiota that differ from that of their unaffected twins who never experienced the lymphoma or its treatment. Further studies are necessary to determine the biological significance of these differences and whether they affect survivorship quality of life.