0 citations0 references

Differential T‐cell subset representation in cutaneous squamous cell carcinoma arising in immunosuppressed versus immunocompetent individuals

Experimental Dermatology2015Vol. 25(3), pp. 245–247

Citations Over Time

Laurence Feldmeyer, Grace Ching, Harina Vin, Wencai Ma, Varun Bansal, Vida Chitsazzadeh, Richard R. Jahan‐Tigh, Emily Y. Chu, Peter J. Fuller, Sourindra N. Maiti, R. Eric Davis, Laurence J.N. Cooper, Kenneth Y. Tsai

Abstract

Cutaneous squamous cell carcinoma (cuSCC) is the 2nd most common skin cancer, of which there are over 700 000 cases per year in the United States, accounting for up to 8000 deaths annually 1. cuSCC is well known to be dramatically affected by the immune status of patients 2. The elevated risk of cuSCC in immunosuppressed individuals is over 50-fold, in which it represents the most common cancer, particularly in organ transplant recipients (OTR) treated with calcineurin inhibitors 2. The risk of iatrogenic cuSCC is lower with the mTOR inhibitors sirolimus and everolimus 3, but prophylaxis or treatment for immunosuppressed patients at high risk for cuSCC remains an unmet medical need. Groundbreaking work in T-cell therapy and immunotherapy has enabled durable responses in multiple cancers. However, given the fact that these therapies are at present systemic, it is unclear how they could be used to treat cuSCC or other cancers in OTRs without causing rejection. Clues may lie in how tumor-infiltrating T cells differ between cuSCC of immunocompetent versus immunosuppressed individuals, but this has been inconclusive to date. Using in vitro expanded cell cultures, Clark and colleagues found that 50% of infiltrating T cells in cuSCC are FOXP3-expressing skin-resident regulatory T cells (Treg) 4. Additionally, FOXP3+ Tregs increase in number from actinic keratosis to cuSCC 5. In one study, cuSCC in OTRs showed reduced proportions of Treg 6, potentially favouring an antitumor response; however, other authors found an increased numbers of FOXP3+ Treg in cuSCC from OTRs 7. Part of the difficulty in identifying potential targets for intervention in cuSCC have been the considerable barriers to gene expression profiling (GEP) of T cells in situ, given that small diagnostic biopsies are often only available formalin-fixed paraffin-embedded (FFPE) tissue samples. The nCounter system (NanoString) uses a highly sensitive, non-enzymatic, digital optical read-out to enable precise quantification of RNA expression, which we have used previously to quantify the TCR diversity in γδT cells 8. Here, we apply the nCounter to compare T-cell GEP in FFPE samples of cuSCC from immunocompetent versus immunosuppressed individuals. How do T cells differ in inflammatory infiltrates of cuSCC arising in immunocompetent versus immunosuppressed individuals? Ten unstained 5-micron sections of invasive, well-differentiated cuSCC (UT Houston) were deparaffinized in xylene and dehydrated in ethanol. Samples were incubated in 10 mm MES pH 6.5, 0.5% SDS and proteinase K (2 mg/ml) at 55 C for 15 min, inactivated at 80 C for 15 min, cleared by centrifugation and applied to the NanoString. Samples were obtained from 12 immunosuppressed (OTR) and 12 immunocompetent (CTRL) patients. Of these, nine OTR samples and 10 CTRL samples passed quality control requirements for RNA data quality. The basic NanoString protocol has been described previously 8. The digital count data were normalized and differentially expressed genes identified. The hypergeometric distribution test was then applied to Gene Ontology set C7 to identify immune-related pathways altered by these expression changes (Supplemental Methods). In this study, 50 cuSCC samples from patients at the University of Texas Medical School, Houston, and the University of Pennsylvania Departments of Dermatology were assessed using immunohistochemistry (IHC) for CD3(+) tumor-infiltrating lymphocytes (TILs) and expression of ID3, OX-40 and FOXO1. As our goal was to identify differences between T cells in cuSCC of OTR versus CTRL patients using a custom 476-gene panel focused on T cells (Table S1), we normalized all measurements in each sample to the average levels of T-cell markers CD3, CD4, CD8 and LCK, which did not differ across the two groups (Table S2). Forty-eight genes were differentially expressed in cuSCC from OTR versus CTRL patients (Fig. 1; Table S2) in over 90% of all pairwise comparisons (Supplemental Methods). FOXP3 expression was strongly upregulated in OTR samples, as were surface molecules and transcription factors associated with Tregs including IL2R, TGFβ1, ID3, RUNX3 and CXCR4. In addition, markers such as CLA were likewise upregulated suggesting an overrepresentation of skin-resident T cells in OTR. Hypergeometric distribution analysis implicated strong upregulation of gene expression signatures associated with Tregs, in vitro differentiation of Tregs, and T-cell exhaustion (Table S3). Importantly, the data revealed the novel findings of increased FOXO1 and OX40 expression in T-cell infiltrates of cuSCC of immunosuppressed patients. These results were confirmed by IHC in additional samples of cuSCC arising in immunocompetent and immunosuppressed patients. This revealed an increased proportion of FOXO1 and OX40 (TNFRSF4)-expressing lymphocytes in cuSCC from OTR patients (Fig. 2). Our study demonstrates that the NanoString platform can be successfully applied to small FFPE tissues to efficiently interrogate gene expression by T cells in situ in skin. We confirm that signatures of Tregs are prominently represented in cuSCCs of OTRs relative to those in immunocompetent individuals, but also identify a potential novel target for intervention, OX40. Recent studies have demonstrated a role for OX40 in the homeostasis of Tregs 9. Agonistic anti-OX40 antibodies can enhance antitumor immunity in preclinical cancer models 10. Our finding of an increased proportion of OX40-positive CD3+ T cells in the cuSCCs of immunosuppressed patients could potentially explain why these tumors are more aggressive than ones in immunocompetent patients, although OX40 has pleiotropic effects in activated T cells 11. FOXO1 was confirmed by IHC to be more highly expressed in T-cell infiltrates around cuSCC arising in OTRs. This transcription factor is required for full Treg activity 12. In conclusion, we show here that a useful, high-throughput, multiplexed method of interrogating the tumor microenvironment has confirmed upregulation of Treg gene expression signatures in the infiltrates surrounding cuSCC arising in OTRs relative to those in immunocompetent individuals. Confirmatory studies with IHC suggest that manipulation of OX40 and AKT/FOXO1 signalling may be particularly useful for cuSCC arising in OTR. We acknowledge the Swiss Cancer Research Foundation (BIL KFS-3344-02-2014) (L.F.), the American Skin Association (K.Y.T.), NIH 5R03AR059246 (K.Y.T.), and MD Anderson Cancer Center (K.Y.T.). L.F., G.C., H.V., V.B, V.C. and P.F. performed research. K.Y.T. designed research and wrote the manuscript with L.F. R.J.T., E.Y.C., S.M. and L.J.N.C. contributed essential reagents. W.M., K.Y.T. and R.E.D. analysed data. The authors have declared no conflicting interests. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Citations Over Time

Abstract

Related Papers