Background The over-expression of stress keratin 17 (K17) has been identified as an immune evasion mechanism in a human papillomavirus (HPV) infection mouse model.1 K17 protein expression has also shown to be a prognostic factor in several HPV-associated cancer types, including head and neck (HNC) and cervical cancer.2 3 Our studies suggest K17 is inversely correlated with CD8 T cell infiltration and response to immune check-point blockade (ICB) in HNC.4 We investigated the expression of K17 and markers of immune activation and response to ICB in anogenital cancers.
Methods Pre-ICB archival FFPE tissue specimens from anogenital cancer patients undergoing ICB-based therapy at the University of Wisconsin-Madison (UW-Madison, n=15) and University of Alabama-Birmingham (n=13) were stained by immunohistochemistry using a validated K17 monoclonal antibody (Abcam, ab109725). Cases were categorized into K17high vs. K17low, as previously described.4 Study endpoints were progression-free survival (PFS), time to treatment failure (TTF) and overall survival (OS). A tissue microarray (TMA) consisting of archival pretreatment tissue samples from 7 ICB-treated anogenital cancer patients from UW-Madison was subject to multiplex single-cell immunephenotyping on the Akoya Biosciences PhenoCycler-Fusion using a 30-plex antibody panel (table 1) which included a custom K17 antibody compatible with the Phenocycler platform requirements (Novusbio, NBP2–47684). Analysis was performed inQuPath v.0.4.3 using the StarDist (arXiv:1806.03535) nuclear segmentation algorithm and phenotyping using artificial neural network training. Correlations between the expression of markers and clinical outcomes were assessed using the Spearman’s coefficient, independent t-test and log rank test.
Results Altogether, 28 patients were included in this study (table 2). Fifteen tumors (53.6%) had K17high expression, and 13 tumors (46.4%) had K17low expression (figure 1). K17 status was significantly correlated with TTF (p=0.03), but not PFS or OS (figure 2A). Among patients receiving pembrolizumab-based therapy (n=21), there were 12 (57.1%) K17high vs. 9 (42.9%) K17low tumors. K17 status was again associated with TTF (p=0.007), but not PFS or OS (figure 2B).
Altogether, single cell immunephenotyping data from 11 TMA cores from 7 patients (table 3) revealed OS at 6 months was significantly correlated with K17 expression (K17+PanCK+/PanCK+, p=0.032), CD68 (CD68+PanCK-/PanCK-, p=0.016) and PD-L1 on tumor cells (PDL1+PanCK+/all cells, p<0.001), (figure 3). There was no correlation between K17 and individual markers.
Conclusions Our findings suggest an inverse trend between K17 expression and clinical outcomes, pending further validation in an expanded patient cohort.
W Wang, et al. ‘Stress keratin 17 enhances papillomavirus infection-induced disease by downregulating T cell recruitment,’ PLoS Pathog, Jan. 2020;16(1):e1008206, doi: 10.1371/journal.ppat.1008206.
L F Escobar-Hoyos, et al. ‘Keratin 17 in premalignant and malignant squamous lesions of the cervix: Proteomic discovery and immunohistochemical validation as a diagnostic and prognostic biomarker,’ Modern Pathology, 2014;27(4):621–630, doi: 10.1038/modpathol.2013.166.
E Regenbogen, et al. ‘Elevated expression of keratin 17 in oropharyngeal squamous cell carcinoma is associated with decreased survival,’ Head and Neck, 2018;40(8):1788–1798, doi: 10.1002/hed.25164.
L P Wang, Lozar T, Golfinos AE, Lee D, Gronski E, Ward-Shaw E, Hayes M, Bruce JY, Kimple RJ, Hu R, Harari PM, Xu J, Keske A, Sondel PM, Fitzpatrick MB, Dinh HQ, ‘Stress Keratin 17 Expression in Head and Neck Cancer Contributes to Immune Evasion and Resistance to Immune-Checkpoint Blockade,’ Clinical Cancer Research, 2022;28(13):2953–2968.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.