High Prevalence of Beta-Human Papilloma Virus in Patients with Concurrent Chronic Lymphocytic Leukemia and Non-Melanoma Skin Cancer

Introduction Patients with chronic lymphocytic leukemia (CLL) are at least twice as likely to develop a second malignancy as their age-matched peers, which is likely secondary chronic CLL-related immunosuppression (Royle 2011). Non-melanoma skin cancers (NMSC) are the most common type of second cancer and those with CLL are seven times more likely to have recurrent NMSC than those without CLL (Mehrany 2005). The etiology of NMSC is likely multifactorial, but we hypothesized that viral load of beta-subtype of human papillomavirus (B-HPV) is correlated with likelihood of NMSC in patients with CLL. Virtually all of the currently available commercial HPV testing targets alpha-subtypes of HPV, but B-HPV has recently been associated with NMSC's (Rollison 2021). The prevalence and viral load of B-HPV in patients with CLL has not been studied.

Methods We collected blood samples from patients with CLL and non-CLL controls (patients with solid malignancies) through the ORIEN AVATAR program and performed whole exome sequencing. We used Kraken2 metagenomic software to identify co-purifying HPV DNA using a custom-built viral database composed of unique sequences from alpha and beta HPV reference genomes. Comparison of viral load between patients with and without CLL, and CLL patients with and without NMSC was performed by a 2 sample t-test. Statistic significance was set as p<0.05.

Results We measured B-HPV viral loads in the peripheral blood of 333 CLL patients and 69 non-CLL control patients. We determined that the average B-HPV viral load in the peripheral blood of CLL patients was double the load in our non-CLL controls (2724.9 versus 1315.0 viral counts; p<0.001). When we classified the B-HPV subtypes, we found higher viral loads of the subtypes 8, 15, 110, 143, 150 in CLL patients than non-CLL controls (p<0.001 for all CLL vs control comparisons). B-HPV subtype 8 (B-HPV8) was the subtype with the highest viral load as determined by exact matches to our Kraken2 HPV database. We identified 29 CLL patients with known diagnosis of NMSC. In this population, the viral load of B-HPV8 was >6-fold higher than that found in the non-CLL control patients (1009.8 vs 149.6 viral sequences; p<0.001; 2-sample t-test).

Conclusions These preliminary data support our hypothesis that B-HPV infection is associated with NMSC in CLL patients. At minimum, it may serve as a biomarker for level of immunosuppression, which predisposes to NMSC. We are currently enrolling a prospective clinical trial for CLL patients with recurrent NMSC (NCT04844528). We plan to sequence both the peripheral blood and the skin cancer samples from these patients to determine whether there is a correlation between B-HPV subtype and NMSC occurrence. We will also evaluate components of the immune system such as T-cell function in an attempt to quantify and correlate immune dysfunction in these patients.

Stephens:Newave: Research Funding; Epizyme: Consultancy; TG Therapeutics: Consultancy; Lilly: Consultancy; Genentech: Consultancy; Celgene: Consultancy; CSL Behring: Consultancy; Beigene: Consultancy; AstraZeneca: Consultancy; AbbVie: Consultancy; Novartis: Research Funding; Mingsight: Research Funding; Karyopharm: Research Funding; JUNO: Research Funding; Acerta: Research Funding; Arqule: Research Funding.