We thank P. Goon and H. Sudhoff for their erudite feedback on our Review (Lechner, M. et al. HPV-associated oropharyngeal cancer: epidemiology, molecular biology and clinical management. Nat. Rev. Clin. Oncol., https://doi.org/10.1038/s41571-022-00603-7; 2022)1, which raises some interesting points of discussion (Goon, P. & Sudhoff, H. HPV-associated oropharyngeal cancer — discussion points. Nat. Rev. Clin. Oncol., https://doi.org/10.1038/s41571-022-00626-0; 2022)2.

Regarding the first point (the existence of a defined premalignant phase in HPV-associated oropharyngeal carcinoma), we do not disagree with this assertion. The legend of figure 2 of our Review1 refers to the much larger body of evidence describing a premalignant phase in cervical cancer, as opposed to the smaller research base for oropharyngeal cancer. We agree that identifying premalignant cells and carcinoma in situ located in the vicinity of diagnosed HPV+ oropharyngeal cancers is possible, as demonstrated in various publications (including our previous publication with P. Goon and H. Sudhoff3); crucially, however, no screening technology or method that enables the routine detection of such lesions in this context is currently available. This lack of a reliable and broadly applicable screening method remains one of the most pressing unmet clinical needs in this field.

The second issue raised indicates the need for further discussion regarding high-risk subgroups of patients with HPV-associated oropharyngeal carcinoma, such as those with a history of tobacco and/or betel nut exposure. Although we do refer to the additional disease-specific mortality caused by smoking, and the occurrence of TP53 and smoking-associated KRAS mutations of the type seen in lung adenocarcinoma, a more extensive description of molecular changes in these subgroups was beyond the scope of our Review1, owing to both word-count restrictions and our remit to focus specifically on the oropharynx. We note that the interesting whole-exome sequencing analysis performed in 2013 (ref.4) mainly refers to oral cavity carcinoma, which is again beyond the intended scope of our Review1. This and other larger whole-exome sequencing studies5 have defined key head and neck squamous cell carcinoma (HNSCC) driver genes, yet analysis of the entire genomic landscape (whole-genome sequencing) of HNSCC is thus far restricted to the study of 150 HNSCCs (103 of which were HPV+) from Gillison and colleagues6, which we discussed in some detail in our Review1. To this end, the ongoing UK 100K genome project includes a subset of ~300 head and neck cancers, which will be analysed using deep (>100×) whole-genome sequencing (I. Reddin et al., unpublished). We hope that large-scale genomic analyses of HNSCC such as these, in combination with further molecular profiling of potential precursor lesions such as those described previously3, will inform the development of much-needed early detection strategies.