Large Consortium Study Reveals More Details on the Link Between Viruses and Cancer

The idea of viruses playing a role in oncogenesis isn’t new. The role human papillomavirus (HPV) plays in cervical cancer and head and neck cancer — and the impact hepatitis B virus (HBV) has on the development of liver cancer — are well documented.¹ But, what about other cancers? Do viruses influence the growth of malignancies, such as renal cell carcinoma (RCC) or blood cancers?

Researchers from the Pan-Cancer Analysis of Whole Genomes (PCAWG) Consortium — equipped with an aggregate of whole-genome sequencing data (WGS) and (in a subset) whole-transcriptome sequencing data (RNA-seq) from tumor tissue samples — endeavored to identify the presence of viral sequences among the various tissue types.¹

These data were pooled from the International Cancer Genomic Consortium (ICGC) and The Cancer Genome Atlas (TCGA) program. In total, there were WGS data from 5354 cancer–normal samples across 38 cancer types, and 1057 tumor RNA-seq data across 25 cancer types.¹ “Between the 2 collections,” said Igor Ristić, PhD, on behalf of the National Cancer Institute’s (NCI’s) Center for Cancer Genomics, “the researchers had access to 2778 cancer samples taken from 2658 donors.” NCI jointly manages the ICGC and TCGA with the National Human Genome Research Institute.

Using high-accuracy pipelines, the PCAWG Consortium reanalyzed the sequencing data to align the results with the human genome and to test for germline and somatic mutations. Then, equipped with the PCAWG Consortium consensus cohort, a PCAWG working group utilized 3 novel, independent, pathogen-discovery pipelines (Computational Pathogen Sequence Identification, Pathogen Discovery Pipeline, and Searching for Pathogens) to detect and quantify viral sequences.¹

To improve consistency among the 3 pipelines, they calculated the relative abundance of virus (viral reads per million extracted reads) at the genus level. If the same virus was detected by 2 pipelines, that was considered a consensus hit.¹ Further, the study authors explained, “We extensively characterized the known and novel viral associations by integrating driver mutations, mutational signatures, gene expression profiles and patient survival data of the same set of tumors analyzed by the PCAWG Consortium.”¹

In all, the researchers found 389 distinct virus–tumor pairs from the data and detected 23 genera among 13% of people with cancer.¹ A total of 85% of the consensus-tumor hits originated from just 5 genera.¹ And among those 5 genera, the results for 3 corroborated existing data, including the role of lymphocryptovirus (eg, Epstein-Barr virus) in gastrointestinal cancer, orthohepadnavirus (eg, HBV) in liver cancer, and alphapapillomavirus (eg, HPV) in head and neck and cervical cancers.¹

Among the more novel information uncovered in this study was the potential role human endogenous retroviruses (HERVs) play in the pathophysiology of certain cancers. More specifically, the endogenous retrovirus (ERV1) expression of HERV was found to be strongest in chronic lymphocytic leukemia when compared with adjoining normal tissue and other tumor types.¹ In addition, ERV1 expression was linked to significantly worse survival outcomes in kidney cancer.¹

These latter results add to a growing body of evidence linking ERV and RCC, said Shifeng Mao, MD, PhD, an attending physician of medical oncology/hematology and chair of Genitourinary Oncology Disease Site Leadership at the Allegheny Health Network Cancer Institute in Pittsburgh, Pennsylvania. “There have been several studies indicating the link between HERV — the remnant of ancient genomic integration of retrovirus — and human cancer. The expression of HERV results in a proinflammatory tumor microenvironment and immune checkpoint activation. Therefore, abnormal expression of certain ERV is associated with response to immune checkpoint blockade.” He added, “There has been increasing interest of HERV in RCC in recent years.”