IARC 60th Anniversary - 19-21 May 2026
Session : 19/05/26 - Posters
Investigating the repurposing potential of PD-1 and PD-L1 inhibitors to expanded cancer patient populations using Mendelian randomisation
BATE T. 1,2, MARTIN R. 2,3, YARMOLINSKY J. 1,2,4, HAYCOCK P. 1,2, DONG M. 5, SMITH-BYRNE K. 6, ATKINS J. 6, BROWN M. 7, HUNG R. 8, LIU G. 7, XU W. 5, ILES M. 9, LAW M. 10, SEVIIRI M. 10, GUSEV A. 11, LIU Y. 11, CONSORTIUM I. 12, CONSORTIUM P. 13
1 Medical Research Council Integrative Epidemiology Unit, University of Bristol, Bristol, United Kingdom; 2 Population Health Sciences, Bristol Medical School, University of Bristol, Bristol, United Kingdom; 3 University Hospitals Bristol and Weston NHS Foundation Trust and the University of Bristol, NIHR Bristol Biomedical Research Centre, Bristol, United Kingdom; 4 Department of Epidemiology and Biostatistics, School of Public Health, Imperial College London, London, United Kingdom; 5 Dalla Lana School of Public Health, University of Toronto, Ontario, Canada; 6 Cancer Epidemiology Unit, Nuffield Department of Population Health, University of Oxford, Oxford, United Kingdom; 7 Princess Margaret Cancer Centre, University of Toronto, Ontario, Canada; 8 Lunenfeld-Tanenbaum Research Institute of Sinai Health, Toronto, Canada; 9 Leeds Institute for Data Analytics, University of Leeds, Leeds, United Kingdom; 10 Statistical Genetics Lab, QIMR Berghofer Medical Research Institute, Brisbane, Australia; 11 Dana-Farber Cancer Institute, Boston, United States; 12 International Lung Cancer Consortium (ILCCO), Lyon, France; 13 Prostate Cancer Association Group to Investigate Cancer Associated Alterations in the Genome (PRACTICAL) consortium, London, United Kingdom
Background: Programmed cell death protein (PD-1) and its ligand programmed death ligand 1 (PD-L1) are immune checkpoint proteins involved in regulating immune responses through T cell activation suppression. This pathway can be exploited by tumour PD-L1 upregulation, thus PD-1 and PD-L1 inhibitors are currently approved for treatment of cancers at different sites (e.g. lung, colorectum). As these medications target a hallmark of cancer, evading host immune responses, we hypothesised that these indications could be expanded to broader cancer patient populations.
Objectives: To assess the repurposing potential of PD-1 and PD-L1 inhibitors, we estimated the genetically proxied effects of PD-1 or PD-L1 plasma protein lowering on risk of mortality for individuals with a breast, colorectal, lung, melanoma, ovarian or prostate cancer diagnosis using Mendelian randomisation (MR).
Methods: Germline genetic instruments were constructed to proxy PD-1 or PD-L1 plasma protein levels. Summary-level MR was performed for breast, colorectal, lung, melanoma, ovarian or prostate cancer survival outcomes using data on 203,031 cancer patients. To investigate risk of collider bias, summary-level MR was performed for cancer incidence outcomes at included sites. The validity of constructed instruments for application to biologically relevant populations (i.e. cancer patients) and tissues (PD-1: T cells, PD-L1: tumour cells) were assessed. Cancer survival for sites with current PD-1 or PD-L1 inhibitor indications (breast, colorectum, melanoma, lung) were used as positive control outcomes.
Results: We observed evidence (p<0.05) to support protective effects of PD-1 or PD-L1 inhibition on risk of mortality for colorectal (HR [95% CI] per SD decrease genetically proxied PD-1: 0.84 [0.71-1.00]; PD-L1: 0.90 [0.81-0.99]) and ovarian (HR [95% CI] per SD decrease genetically proxied PD-1: 0.88 [0.78-0.99]) cancer patients. Meta-analysis of findings across included sites suggested stronger evidence for an effect of PD-1 (HR [95% CI]=0.91 [0.85-0.96]) than PD-L1 (HR [95% CI]=1.00 [0.96-1.03]). There was no strong evidence (p<0.05) to support effects of PD-1 or PD-L1 on cancer incidence for included sites, suggesting low risk of collider bias impacting our analyses. The lead instruments were associated with respective plasma protein levels in concordant directions of effect for UK Biobank populations both when restricted to or unrestricted to participants with cancer diagnoses. However, there was no strong evidence (false discovery rate p<0.05) to support associations of the lead instruments with expression of respective protein-coding genes in relevant cell types.
Conclusions/Implications: There was no strong evidence to support effects of PD-1 or PD-L1 inhibition on survival for patients diagnosed with cancers at the majority of included sites. However, there was some evidence to support PD-1 inhibitor repurposing to ovarian cancer treatment, and PD-1 or PD-L1 inhibitor repurposing to broader colorectal cancer patient populations than current approvals. ?Our constructed instruments were validated with respect to the biologically relevant population, yet applicability for proxying mechanisms at the biologically relevant tissues were less clear. Therefore, findings should be interpreted with caution due to highlighted potential violations of the MR instrument relevance assumption, lack of clear evidence for effects on survival at sites with current indications (i.e. positive control outcomes) and low statistical power.