Fisher's exact test.

### **4. Discussion**

The incidence of HPV-related HNSCC is currently increasing and gaining importance, while tumors associated to tobacco and alcohol consumption are declining [4,5]. Recent studies have established important differences between HNSCC depending on HPV infection status, including clinical, biological, and molecular features, emphasizing the presence of less genetic alterations, a better response to chemotherapy and radiotherapy, and a more favorable prognosis for HPV-related HNSCC [2,4,5]. These differences could change the way we diagnose, treat, and manage HNSCC. One of the most frequent genetic alterations found in HNSCC is the amplification of the 11q13 locus, which has been associated with increased tumor growth, proliferation, and dissemination [8,9].

Given the importance of the 11q13 locus in HNSCC and to contribute to the molecular characterization of these tumors, we conducted a study on a large unbiased cohort of 392 homogeneous surgically treated HNSCC patients to investigate the role of 11q13 amplification in relation to HPV status. This was accomplished by assessing the specific relationship of the protein expression and amplification of the *CTTN, CCND1*, and *ANO1* genes mapping within this locus. Immunohistochemical analysis of CCND1, CTTN, and ANO1 revealed that the expression of these three proteins was strongly and inversely correlated with HPV infection. Noteworthy, all the HPV-positive cases showed negative ANO1 expression, and 28 out of 30 HPV-positive cases had also negative CTTN expression. Likewise, the analysis of gene copy amplification by real-time PCR consistently showed that amplifications of *CTTN*, *CCND1* and *ANO1* were frequently detected in HPV-negative tumors (ranging from 42 to 61%), while absent or rare in HPV-positive tumors (0–15%). Mechanistically, we found that the contribution of gene amplification to protein expression varied widely depending on each gene. Even though tumors harboring amplification also concomitantly expressed high protein levels, positive CCND1 and CTTN expression occurred at a higher frequency than gene amplification, while ANO1 protein expression was less frequent than *ANO1* gene amplification. Additional regulatory mechanisms (transcriptional and post-translational) should contribute to protein expression, as previously reported [18]. A limitation of our study is the use of tissue microarrays to evaluate protein expression, which may constitute a drawback to assess the possible influence of tumor heterogeneity. To minimize this, three different representative tumor areas were selected from each tissue block and analyzed in the TMAs. Of note, the results showed that these proteins presented homogeneous and highly concordant expression patterns in the three tissue punches from each tumor.

These results were further and significantly validated using an independent cohort of 279 TCGA HNSCC patients, thus confirming that the mRNA expression levels of CCND1, CTTN, and ANO1 were significantly increased in a high proportion of HPV-negative tumors, whereas most HPV-positive patients exhibited normal mRNA levels of all these genes. Similar observations were obtained by analyzing the frequencies of copy number alterations. These findings are also consistent with some studies that reported that 11q13 amplifications and gains were less frequent in HPV-related tumors [2,4]. This is also in agreement with the assumption made by Kostareli *et al.* [6] that, in HPV-positive tumors, a lower amount of genetic alterations is required to achieve the malignant phenotype, because of the inactivation of p53 and pRb proteins by the viral E6 and E7 oncoproteins. To our knowledge, this is the first study to assess specifically the protein expression and copy number alterations of various genes mapping at the 11q13 amplicon in relation to HPV infection status using two large independent cohorts of HNSCC patients. Our study unveils important differences regarding the expression and amplification of the *CCND1*, *CTTN* and *ANO1* genes between HPV-related and unrelated tumors. It is worth mentioning that the two HNSCC cohorts selected for study are representative, sharing multiple of the unique characteristics reported for HPV-positive tumors [13], such as a clear prevalence in the oropharynx, low tobacco and alcohol consumption by the patients, lower tumor stage, basaloid histological pattern, a younger patients' age at diagnosis, lower mutations and CNA, and above all, a better prognosis. A recent study by Dixit *et al.* [21] provided the first evidence of a link between ANO1 expression and gene amplification and HPV status using a series of 64 pharyngeal tumors and tissue microarrays for IHC evaluation. Therefore, our results further and significantly strength and

validate these preliminary findings on ANO1 protein expression using a large independent cohort of 392 HNSCC patients, as well as on ANO1 mRNA levels in the TCGA cohort of 530 HNSCC patients.

Together, these findings suggest that the *CCND1*, *CTTN* and *ANO1* genes within the 11q13 amplicon could play a significant role in HPV-negative tumors but not in HPV-positive tumors. Given that 11q13 amplification has been associated with poor prognosis in HNSCC patients, and, in particular, these three genes have been involved in tumor progression and resistance to radio-, chemotherapy [11,22–26], and EGFR-targeted therapies [27], the herein found distinctive molecular alterations presumably could contribute to the clinical and biological differences between these two different HNSCC subtypes and explain the better prognosis and response to radiotherapy and chemotherapy associated to HPV-positive tumors. In fact, we proved the impact of these molecular alterations on patient survival. In particular, CTTN and ANO1 overexpression, rather than gene amplification, was more frequent and found to associate with a worse clinical outcome in two TCGA cohorts of 279 and 530 HNSCC patients. Nevertheless, given the size of the 11q13 amplicon, it cannot be ruled out that the overexpression of these genes may be secondary and that other genes within the 11q13 amplicon could act as true drivers of HNSCC progression. A recent study has also identified four genes (*ORAOV1*, *CPT1A*, *SHANK2* and *PPFIA1*) as important drivers of 11q13 amplification that showed an impact on prognosis [28]. Therefore, various genes within the 11q13 amplicon could cooperatively contribute to the differences in prognosis and clinical outcome between HPV-positive and HPV-negative tumors. In summary, various studies including ours have provided strong evidence demonstrating that HPV-related and unrelated tumors are two different entities; hence, we consider that the management of HNSCC patients should move toward a more personalized approach.

### **5. Conclusions**

*CTTN*, *CCND1* and *ANO1* amplification and overexpression were frequent in HPV-negative tumors and correlated with reduced patient survival, while absent or very rare in HPV-positive tumors. Therefore, these molecular alterations could contribute to the distinct clinical outcomes of these two HNSCC entities and serve as the basis to design more personalized therapeutic strategies for HNSCC patients.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2077-0383/7/12/501/s1, Figure S1. Schematic representation of the experimental setup designed to analyze CCND1, ANO1 and CTTN protein expression and gene amplification in relation to HPV status in a cohort of 392 HNSCC patients, Figure S2. Clinical and molecular features of the TGCA cohort of 279 HNSCC patients, according to the HPV status. Schematic representation showing the HPV incidence by HNSCC subsites, mutation frequencies, and Kaplan–Meier survival curves (*p* = 0.024, Log-rank test). Raw data of CTTN, CCND1 and ANO1 protein scores in relation to the HPV status for the total cohort of 392 HNSCC patients, Table S1: Clinicopathologic features of the validation cohort of 279 HNSCC patients from the TCGA.

**Author Contributions:** Conceptualization, J.P.R. and J.M.G.-P.; Funding acquisition, X.L., J.P.R. and J.M.G-P; Investigation, F.H.-P., S.T.M., P.A.-A., R.G.-D, M.Á.V., L.A., N.d-R.-I., R.R., J.P.R. and J.M.G-P; Methodology, F.H.-P., S.T.M., E.A., L.A.-D. and A.A.; Resources, X.L., L.A., M.M., A.A. and J.P.R.; Supervision, J.M.G.-P.; Validation, F.H.-P., S.T.M. and S.Á.-T.; Visualization, F.H.-P., P.A.-A. and J.M.G.-P.; Writing—original draft, P.A.-A. and J.M.G.-P.; Writing—review & editing, F.H.-P., S.T.M., X.L., L.A., M.M, R.R. and J.P.R.

**Funding:** This study was supported by grants from the Plan Nacional de I+D+I 2013-2016 ISCIII (PI13/00259), RD12/0036/0015 of Red Temática de Investigación Cooperativa en Cáncer (RTICC), PI16/00280, and CIBERONC (CB16/12/00390 and CB16/12/00401), CIBER-BBN (CB06/01/1031) Spain, Fundación Merck Salud (17-CC-008), the Instituto de Investigación Sanitaria del Principado de Asturias (ISPA), PCTI-Asturias (GRUPIN14-003), Fundación Bancaria Caja de Ahorros de Asturias-IUOPA, and the FEDER Funding Program from the European Union. STM (Sara Borrell Program-CD16/00103) and RR (Miguel Servet II Program-CPII16/00049) were recipients of fellowships from ISCIII.

**Acknowledgments:** We thank the samples and technical assistance kindly provided by the Principado de Asturias BioBank (PT13/0010/0046), financed jointly by Servicio de Salud del Principado de Asturias, Instituto de Salud Carlos III and Fundación Bancaria Cajastur and integrated in the Spanish National Biobanks Network, and IIB Sant Pau. We also thank Aitana Vallina for excellent technical assistance, Pablo Martínez-Camblor for his assistance with statistical analyses.

**Conflicts of Interest:** The authors declare no conflict of interest. Institutional support to L.A. and M.M.: HPV epidemiological studies sponsored by GlaxoSmithKline, Merck and Seegene.

### **References**


© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
