WO2019134994A1

WO2019134994A1 - Prognostic biomarkers for human papillomavirus positive cancers

Info

Publication number: WO2019134994A1
Application number: PCT/EP2019/050239
Authority: WO
Inventors: Christian Gaiddon; Alain JUNG
Original assignee: Universite De Strasbourg; INSERM (Institut National de la Santé et de la Recherche Médicale); Centre Paul Strauss
Priority date: 2018-01-08
Filing date: 2019-01-07
Publication date: 2019-07-11

Abstract

The present invention relates to new biomarkers for predicting the clinical outcome of a subject having a Human papillomavirus (HPV) positive cancer.

Description

Prognostic biomarkers for Human Papillomavirus positive cancers

FIELD OF THE INVENTION

The present invention relates to the fields of genetics and medicine, in particular of oncology. It provides new prognostic biomarkers for Human Papillomavirus positive cancers.

BACKGROUND OF THE INVENTION

Human papillomavirus (HPV) infection is now a well-established cause of cancer and there is growing evidence of HPV being a relevant factor in cervical and anogenital cancers (anus, vulva, vagina and penis) as well as in head and neck cancers. The majority of head and neck cancers are associated with high tobacco and alcohol consumption. However, increasing trends in the incidence at specific sites suggest that other etiological factors are involved, and infection by certain high-risk types of HPV have been reported to be associated with head and neck cancers, in particular with oropharyngeal cancer.

Cancers induced by HPV infection are a distinct clinical subgroup of cancers with pathological and molecular features that are different from their HPV-negative counterpart. Patients with HPV-positive cancer also have a prolonged disease-free and overall survival but can display a rate of distant metastases that is similar to the one observed in HPV-negative patients. Intriguingly, these metastases occur in atypical anatomic localizations with a higher rate of metastasis to the brain, compared to the metastatic spread spectrum that is usually observed in HPV-negative cancer, suggesting that HPV-positive cancer also constitute a heterogeneous group of malignancies.

There is thus a need to distinguish HPV⁺ cancers with high or low probability of distant metastasis linked to a poor prognosis and survival chances.

HPV⁺ tumours are also thought to be more sensitive to both chemo- and radio- therapies. However, these protocols are associated with marked acute and late toxicities. It has therefore been proposed that alternative therapeutic protocols could be used for the management of patients with HPV-positive cancer in order to spare them high-grade toxicities induced by aggressive radiation and chemo-radiation protocols, and to improve the therapeutic index.

There is thus a need to identify means to predict the clinical outcome of a subject having HPV⁺ cancer. This would allow to select patients affected with HPV positive cancer with a good or poor prognosis and to determine whether these patients are susceptible to benefit or not from a therapeutic de-escalation. The present invention seeks to meet these and other needs. SUMMARY OF THE INVENTION

The present invention relates to the use of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63 for predicting the clinical outcome of a subject having a Human papillomavirus (HPV) positive cancer.

Accordingly, the present invention relates to an in vitro method for predicting the clinical outcome of a subject having a HPV positive cancer, wherein the method comprises: detecting at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4 and DNr63 in a cancer sample from said subject, and

determining the expression level of said at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63 in said cancer sample, the expression level being indicative of the clinical outcome.

The expression level of the biomarker can be determined by measuring the quantity of the mRNA transcripts, for instance by quantitative RT-PCR, real time quantitative RT-PCR or RNA-seq. The expression level of the biomarker can also be determined by measuring the quantity of proteins, for instance by immunohistochemistry.

The cancer can be selected from the group consisting of head and neck cancers, oropharyngeal cancers, hypopharynx cancers, oesophageal cancers, cervical cancer, anal cancer, vaginal cancers, vulvar cancers, penile cancers, endometrial cancers and uterine cancers. Particularly, the oropharyngeal cancers are throat, tongue or tonsils cancers, preferably oropharyngeal squamous cell carcinoma.

In one embodiment, the expression level of the biomarker is combined with the gender and/or age of the subject, tobacco smoking, tumour stage and/or size for prognosis determination.

For instance, high level of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, KRT6B and/or DNr63 and/or low level of THBS4 are predictive of a good prognosis, and wherein a good prognosis is preferably a good survival prognosis, a good metastasis-free survival, a decrease disease recurrence and/or a decrease metastasis occurrence. Particularly, a good prognosis is a decrease of distant metastasis occurrence.

In a particular embodiment, the biomarker is selected from the group consisting of S100A9, THBS4 and DNr63, alone or in combination. More particularly, the biomarkers are S100A9 and

THBS4.

In addition, the invention relates to a method for selecting a subject susceptible to benefit from a therapeutic de-escalation that comprises: determining the clinical outcome of a subject having a Human papillomavirus positive cancer by the method and

selecting a subject with a good prognosis as susceptible to benefit from a therapeutic de-escalation.

Also, the invention relates to a method for selecting a subject susceptible to better respond to an immunotherapy that comprises: determining the clinical outcome of a subject having a Human papillomavirus positive cancer by the method and

selecting a subject with a good prognosis as susceptible to better respond to an immunotherapy.

Particularly, in such methods the biomarkers are S100A9 and THBS4, where low level of S100A9 and high level of THBS4 are predictive of a poor diagnosis, which can be a poor survival prognosis, an early disease progression, a lymph node involvement, an increased disease recurrence, especially after resection and/or treatment, or more preferably an increased metastasis occurrence.

Finally, the present invention relates to the use of a kit comprising means for measuring the expression level of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, and DNr63 for (i) predicting the clinical outcome of a subject having a HPV positive cancer (ii) selecting a subject affected with HPV positive cancer with a good or poor prognosis and/or (iv) determining whether a subject affected with a HPV positive cancer is susceptible to benefit from a therapeutic de-escalation or from an immunotherapy.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1: FIG 1A. Principal Component Analysis (PCA) of the transcriptomic data of 8 HPV-positive OSCC. This analysis highlights two distinct subgroups of tumours (Cluster 1 in black; cluster 2 in red).

FIG IB. Hierarchical clustering and heatmap analysis of the genes that are differentially expressed between tumour samples in cluster 1 (black) and cluster 2 (red).

FIG 1C. Kaplan-Meier analysis of the overall survival of patients in cluster 1 (black) and cluster 2 (red). Differences were found to be statistically significant (p=0,004).

Figure 2: Correlation matrix of the expression of the TP63 gene and the genes known to be regulated by DNr63 (Barbieri et al. Cancer Res 2006) based on the transcriptomic data generated from HPV-positive OSCC. A scale with the Pearson Correlation Coefficient is shown. Correlations between the expression of TP63 and genes of interest appear as blue dots in the matrix; anti correlations between the expression of TP63 and genes of interest appear as red dots. Genes previously shown to be downregulated when DNr63 is knocked-down are indicated with a green dot (left to the gene), and genes known to be up-regulated upon DNr63 inhibition are indicated with a red dot (left to the gene gene).

Figure 3: Immunohistochemical analysis of the expression of the DNr63, S100A7, S100A9, KRT6B and THBS4 proteins in serial slides from formalin-fixed paraffin-embedded specimen of HPV- positive OSCC. Tumours were stratified according to the expression of DNr63, in ANp63-high (middle panels) and ANp63-low (right panels) expressing OSCC. Positively-labelled normal epithelium is shown in left panels.

Figure 4: Gene expression analysis of S100A7, S100A9, KRT6B and THBS4 in HPV-positive OSCC stratified according to the expression of the DNr63 protein (ANp63-high: dark grey plot boxes; ANp63-low/neg: light grey plot boxes). Expression levels in ANp63-high and ANp63-low/neg OSCC were compared and were found to be statistically different.

Figure 5: FIG 5A. Kaplan-Meier analysis of the metastasis-free survival of HPV-positive patients stratified according to the DNr63 protein expression level (ANp63-high vs. ANp63-neg/low, as determined by IHC). Differences were found to be statistically significant (log-rank test).

FIG 5B. Kaplan-Meier analysis of the metastasis-free survival of HPV-positive patients stratified according to the expression of DNr63 protein expression level (ANp63-high vs. ANp63-neg/low, as determined by IHC), the history of tobacco smoking and the N stage (non-smoker; N0-N2a vs. current/previous smoker; N2b-N3) Differences were found to be statistically significant (log-rank test).

FIG 5C. Kaplan-Meier analysis of the metastasis-free survival of HPV-positive patients stratified according to the combined expression of the THBS4 and S100A9 genes. Differences were found to be statistically significant (log-rank test).

Figure 6: Kaplan-Meier analysis of the metastasis-free survival of HPV-positive patients stratified according to the expression of S100A7 (A), S100A9 (B) and THBS4 (C). Patients were stratified in high and neg-low expressing tumours, as determined by IHC. Differences were found to be statistically significant (log-rank test).

Figure 7: Kaplan-Meier analysis of the metastasis-free survival of HPV-positive patients stratified according to the expression of the S100A7 (A), S100A9 (B), SERPINB1 (C) KRT6B (D), SPRR1A (E), THBS4 (F) and DNr63 (F) genes. Patients were stratified in positive (+) and negative (-) tumours as determined by RT-qPCR. Differences were found to be statistically significant (log-rank test). DETAILED DESCRIPTION OF THE INVENTION

The inventors demonstrate that expression levels of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63 can be used to determine the clinical outcome of a subject having a Human papillomavirus positive (HPV⁺) cancer. The prognosis of the clinical outcome can be useful for proposing to the patient the most appropriate treatment, particularly a de-escalation therapy or an immunotherapy.

Definitions

The term "cancer" or "tumour", as used herein, refers to the presence of cells possessing characteristics typical of cancer-causing cells, such as uncontrolled proliferation, immortality, metastatic potential, rapid growth and proliferation rate, and certain characteristic morphological features. This term includes early stage, localized cancer, later stage, locally advanced cancer; and metastatic stage cancer in any type of patient.

The terms "Human papillomavirus positive cancer" or "HPV positive cancer" or "HPV⁺ cancer" are interchangeable and refer to a cancer that is positive for HPV. Determining the HPV status of the cancer comprises determining if a cancer tumour expresses one or more proteins derived from an HPV or a nucleotide sequence encoding the one or more proteins. In certain embodiments, more than at least about 50 percent, at least about 55 percent, at least about 60 percent, at least about 65 percent, at least about 70 percent, at least about 75 percent, at least about 80 percent, at least about 85 percent, at least about 90 percent, at least about 95 percent, at least about 99 percent, or at least about 100 percent of the tumour cells show strong and diffuse nuclear and cytoplasmic staining by an immunohistochemistry against pl6 and the tumours are considered HPV-positive. HPV positivity is defined by the detection of the viral genome by qPCR and particularly trough the detection of the viral RNA that encodes the E6 and E7 oncoproteins. There are more than 200 related viruses in the HPV family, including subtypes 6, 11, 16, 18, 30, 31, 33, 34, 35, 39, 40, 42, 43, 44, 45, 51, 52, 53, 54, 55, 56, 57, 58, 59, 66, 68, and other unidentified subtypes. In some embodiments, the HPV subtype is HPV subtype 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, 68, or any combination thereof. In some embodiments, the HPV subtype is HPV subtype 16. In some embodiments, the HPV subtype is HPV subtype 18. A cancer can be determined to be an HPV-positive cancer by any suitable means, which are well known to the skilled person. In particular, detection of HPV and/or HPV serotyping can be proceeded on the cancer sample by any method known by the skilled person of the art, e.g. RT-PCR with primers targeting HPV genome or antibodies against any viral protein.

The term "cancer sample" or "tumour sample" refers to any sample containing tumour cells derived from a patient or a subject. In particular, tumour cells may be obtained from fluid sample such as blood, plasma, urine and seminal fluid samples as well as from biopsies, organs, tissues or cell samples. In a preferred embodiment, tumour cells are obtained from tumour biopsy or resection sample from the patient. Preferably, the sample contains only tumour cells. Preferably the cancer sample contains nucleic acids and/or proteins. Preferably, cancer samples are HPV positive cancer samples. Optionally, samples containing tumour cells may be treated prior to their use. As example, a tumour cell enrichment sorting may be performed. The sample may be treated prior to its use. It may be fresh, frozen or fixed (e.g. formaldehyde or paraffin fixed) sample.

As used herein, the terms "subject", "individual" or "patient" are interchangeable and refer to an animal, preferably to a mammal, even more preferably to a human. However, the term "subject” can also refer to non-human animals, in particular mammals such as dogs, cats, horses, cows, pigs, sheep and non-human primates, among others.

As used herein, the terms "marker" and "biomarker" are interchangeable and refer to biological parameters that permit the selection of patients who will have a poor or good prognosis and/or will benefit from a specific treatment. This term refers particularly to "tumour biomarkers". It is a measurable indicator for predicting the clinical outcome of a subject having cancer. Hereafter S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B, and DNr63 are considered as biomarkers.

As used herein, the terms "clinical outcome" and "prognosis" are interchangeable and refer to the determination as to whether a subject is likely to be affected by a cancer or cancer recurrence or metastasis, preferably distant metastasis. The skilled person often makes a prognosis on the basis of one or more diagnosis biomarkers, the presence, absence, or amount of which is indicative of cancer recurrence or metastasis.

As used herein, the term "poor prognosis" refers to a decreased patient survival and/or an early disease progression and/or an increased disease recurrence and/or an increased metastasis formation and/or distant metastasis occurrence and/or lymph nodes involvement. Conversely, the term "good prognosis" refers to an increased patient survival and/or a delayed disease progression and/or a decreased disease recurrence and/or a decreased metastasis formation.

The terms "distant metastasis" or "distant cancer" are interchangeable and refer to cancer that has spread from the original or primary tumour to distant organs or lymph nodes.

The term "distant metastasis free survival" refers to a temporal measure from a defined start point e.g. diagnosis, or treatment to the appearance of distant metastasis.

The terms "therapeutic de-escalation" and "de-escalation therapy" are interchangeable and refer to a strategy that is mainly concerned with decreasing the doses or the frequency of a treatment to the patient, decreasing the risk of side effects, or simplifying the procedures or to a strategy with an alternative therapeutic agent, alone or in combination with radiotherapy and/or chemotherapy. In chemotherapy, therapeutic de-escalation can be based on a better assessment of the tumour's profile, i.e. expression of specific biomarkers. In particular, a de-escalation therapy can refer to a chemo- and/or radio-therapy at subtherapeutic dose. As used herein, "subtherapeutic dose" means a dose of a therapeutic compound that is lower than the usual or typical dose of the therapeutic compound when administered alone for the treatment of a cancer.

As used herein, the term "immunotherapy" refers to a cancer therapeutic treatment using the immune system to reject cancer. The therapeutic treatment stimulates the patient's immune system to attack the malignant tumour cells. This includes immunization of the patient with tumoral antigens (e.g., by administering a cancer vaccine), in which case the patient's own immune system is trained to recognize tumour cells as targets to be destroyed, or administration of molecules stimulating the immune system such as cytokines, or administration of therapeutic antibodies as drugs, in which case the patient's immune system is recruited by the therapeutic antibodies to destroy tumour cells. In particular, antibodies are directed against specific antigens such as the unusual antigens that are presented on the surfaces of tumours.

The terms "quantity," "amount," and "level" are used interchangeably herein and may refer to an absolute quantification of a molecule in a sample, or to a relative quantification of a molecule in a sample, i.e., relative to another value such as relative to a reference value as taught herein, or to a range of values for the biomarker. These values or ranges can be obtained from a single patient or from a group of patients.

As used herein, the "reference value" or "cut-off value" are used interchangeably and refer to a threshold value. It can be an absolute value; a relative value; a value that has an upper and/or lower limit; a range of values; an average value; a median value; a mean value; a statistic value; a cut-off or discriminating value; or a value as compared to a particular control or baseline value. Typically, a "threshold value" or "cut-off value" can be determined experimentally, empirically, or theoretically. A threshold value can also be arbitrarily selected based upon the existing experimental and/or clinical conditions, as would be recognized by a person of ordinary skilled in the art. A reference value can be based on an individual sample value, such as a value obtained from a sample from the individual tested but at an earlier point in time, or a value obtained from a sample from a subject other than the individual tested, or a "normal" individual that is an individual or a group of individuals identified as having a healthy status or not diagnosed with any HPV positive cancer and/or metastasis derived from HPV positive cancer.

Preferably, the person skilled in the art may compare the biomarkers activation levels with a defined threshold value. In one embodiment of the present invention, the threshold value is derived from the biomarkers activation level (or ratio, or score) determined in a sample derived from one or more healthy subjects, especially who are not suffering from HPV+ cancer and/or from metastasis derived from HPV positive cancer. Furthermore, retrospective measurement of the biomarkers activation level (or ratio, or scores) in properly banked historical subject samples may be used in establishing these threshold values.

The threshold value has to be determined in order to obtain the optimal sensitivity and specificity according to the function of the test and the benefit/risk balance (clinical consequences of false positive and false negative). Typically, the optimal sensitivity and specificity (and so the threshold value) can be determined using a Receiver Operating Characteristic (ROC) curve based on experimental data.

"S100A9" designates S100-A9 which is a member of the S100 family which include at least 13 members. S100 proteins are localized in the cytoplasm and/or nucleus of a wide range of cells, and involved in the regulation of a number of cellular processes such as cell cycle progression and differentiation. Alternative names are Calgranulin-B (CAGB) , Calprotectin L1H subunit, Leukocyte LI complex heavy chain, Migration inhibitory factor-related protein 14 (MRP-14). SA100A9 is described in databases under the following accession numbers: Gene ID: 6280, UniGene Hs.112405. This protein is disclosed in UniProt under accession number: P06702. The GenBank entry of the sequence of the protein and mRNA are respectively NP_002956.1. and NM_002965.3.

"S100A7" designates S100-A7 which is a member of the S100 family which include at least 13 members. Alternative names are Psoriasin (PSOR1) and S100 calcium-binding protein A7. SA100A7 is described in databases under the following accession numbers: Gene ID: 6278, UniGene Hs.112408. This protein is disclosed in UniProt under accession number: 31151. The GenBank entry of the sequence of the protein and mRNA are respectively: NP_002954.2. and NM_002963.3.

"SERPIN Bl" gene is a member of the serpin family of proteinase inhibitors and encoding Leukocyte elastase inhibitor (LEI). Members of this family maintain homeostasis by neutralizing overexpressed proteinase activity through their function as suicide substrates. Alternative names are Monocyte/neutrophil elastase inhibitor (El or M/NEI), or Peptidase inhibitor 2 (PI-2). SERPIN Bl is described in databases under the following accession numbers: Gene ID: 1992, UniGene Hs.381167. This protein is disclosed in UniProt under accession number: 30740. The GenBank entry of the sequence of the protein and mRNA are respectively: NP_109591.1. and NM_030666.3.

"SPRR1A" or "SPR-IA" gene (Small Proline Rich protein 1A) codes for Cornifin-A. SPRR genes encode proteins that are strongly induced during differentiation of human epidermal keratinocytes in vitro and in vivo and are a class of cornified envelope precursor proteins. An alternative name is 19 kDa pancornulin and SPRK. SPRR1A is described in databases under the following accession numbers: Gene ID: 6698, UniGene Hs.46320. This protein is disclosed in UniProt under accession number: P35321. The GenBank entry of the sequence of the protein and mRNA are respectively: NP_001186757.1. and NM_001199828.1. for variant 1, and NP_005978.2. NM_005987.3 for variant

2. "THBS4" gene is the Thrombospondin-4 gene and belongs to the thrombospondin protein family. Thrombospondin family members are adhesive glycoproteins that mediate cell-to-cell and cell-to-matrix interactions. An alternative name is TSP4. THBS4 is described in databases under the following accession numbers: Gene ID: 7060, UniGene Hs.211426. This protein is disclosed in UniProt under accession number: P35443. The GenBank entry of the sequence of the protein and mRNA are respectively: NP_003239.2. NM_003248.5 for the precursor; NM_001306212.1. NP_001293141.1. for variant 2; NP_001293142.1. NM_001306213.1. for variant 3;

NP_001293143.1. NM_001306214.1. for variant 4.

"KRT6B" gene is the Keratin type II cytoskeletal 6B gene. There are at least six isoforms of human type II keratin-6 (K6). Alternative names are Cytokeratin-6B (CK-6B), Keratin-6B (K6B), Type- II keratin KblO. KRT6B is described in databases under the following accession numbers: Gene ID: 3854, UniGene Hs.708950. This protein is disclosed in UniProt under accession number: P04259. The GenBank entry of the sequence of the protein and mRNA are respectively: NP_005546.2. and NM_005555.3.

"DNr63" is an isoform of TP63 gene that is a member of the p53 family of transcription factors. The functional domains of p53 family proteins include an N-terminal transactivation domain, a central DNA-binding domain and an oligomerization domain. Alternative splicing of this gene and the use of alternative promoters results in multiple transcript variants encoding different isoforms that vary in their functional properties. DNr63 is described in databases under the following accession numbers: Gene ID: 8626, UniGene Hs.137569. This protein is disclosed in UniProt under accession number: Q9H3D4. DNr63 gene or protein could also be DNr63 variant, for example DNr63 alpha (isoform 2 of p53), DNr63 beta (isoform 4 of p53), DNr63 gamma (isoform 6 of p53).

Biomarkers for the prognosis of patients having HPV+ cancers

In a first aspect, the present invention concerns the use of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63 for predicting the clinical outcome of a subject having a Human papillomavirus positive (HPV⁺) cancer.

In a second aspect, the invention concerns an in vitro method for predicting the clinical outcome of a subject having a HPV positive cancer. The method comprises the detection of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63 in a cancer sample from said subject, and/or the determination of the expression level of said at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63 in said cancer sample. The expression level of the biomarkers is indicative of the clinical outcome of the patient, i.e., a good or poor prognosis, e.g. patient survival, cancer recurrence and/or metastasis occurrence, preferably distant metastasis occurrence.

Optionally, the expression level of the biomarker(s) can be combined with the gender and/or age of the subject, his/her history of tobacco smoking, the tumour stage and/or size for prognosis determination.

In a particular embodiment, high level(s) of at least one S100A9, S100A7, SERPINB1, SPRR1A and/or DNr63 and/or low level of THBS4 are predictive of a good prognosis.

A good prognosis is preferably a good survival prognosis, a good metastasis-free survival, a decrease disease recurrence and/or a decrease metastasis occurrence, preferably a decrease of distant metastasis occurrence.

Optionally, the method enables to select patient susceptible to benefit from therapeutic de-escalation or immunotherapy, preferably to select patient with a good prognosis. In this context, the present invention relates to a method for treating a patient having a Human papillomavirus positive (HPV⁺) cancer, comprising determining the clinical outcome of the subject as detailed herein based on the expression level of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, and DNr63 and administering a therapeutic de- escalation treatment or an immunotherapy to the subject if the clinical outcome is a good prognosis.

In another embodiment, low level(s) of at least one S100A9, S100A7, SERPINB1, SPRR1A and/or DNr63 and/or high level of THBS4 are predictive of a poor diagnosis.

A poor prognosis is preferably a poor survival prognosis, an early disease progression, a lymph node involvement, an increased disease recurrence, especially after resection and/or treatment, more preferably an increased metastasis occurrence, even more preferably an increased distant metastasis occurrence.

Preferably, the biomarker(s) for predicting the clinical outcome of a subject having a HPV⁺ cancer can be S1009 and/or S100A7 and/or SERPINB1 and/or SPRR1A and/or THBS4 and/or DNr63.

The biomarkers for predicting the clinical outcome of a subject having a HPV⁺ cancer can be any combination of biomarkers selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63. For instance, such combination includes a combination of 2, 3, 4, 5 or 6 biomarkers selected from this group, such as S100A9 and S100A7; S100A9 and SERPINB1; S100A9 and SPRR1A; S100A9 and THBS4; S100A9 and DNr63; S100A7 and SERPINB1; S100A7 and SPRR1A; S100A7 and THBS4; S100A7 and DNr63; SERPINB1 and SPRR1A; SERPINB1 and THBS4; SERPINB1 and DNr63; SPRR1A and THBS4; SPRR1A and DNr63; THBS4 and DNr63; S100A9, THBS4 and DNr63; S100A7, S100A9, THBS4 and DNr63; KRT6B and S100A9, KRT6B and S100A7, KRT6B and SERPIN1B, KRT6B and SPRR1A, KRT6B and THBS4; KRT6B and DNr63 .

In a particular aspect, the combination includes S100A9 and THBS4 and optionally at least one biomarker selected from the group consisting of S100A7, SERPINB1, SPRR1A, KRT6B and DNr63.

In a particular aspect the combination includes S100A9, THBS4 and S100A7, or SA100A9, THBS4 and SEPRINB1, or SA100A9, THBS4 and SPRR1A or SA1009, THBS4 and DNr63 or SA1009, THBS4 and KRT6B.

In another particular aspect, the combination includes S100A9, THBS4 and DNr63, and optionally at least one biomarker selected from the group consisting of S100A7, SERPINB1, KRT6B and SPRR1A.

In a particular aspect the combination includes S100A9, THBS4, DNr63 and S100A7, or SA100A9, THBS4, DNr63 and SEPRINB1, or SA100A9, THBS4, DNr63 and SPRR1A or S100A9, THBS4, DNr63 and KRT6B.

In a further aspect, the biomarker for predicting the clinical outcome of a subject having a HPV⁺ cancer can be S100A9, THBS4 and/or DNr63, preferably S100A9 and/or THBS4.

In a more preferred embodiment, the present invention concerns at least one biomarker selected from the group consisting of S100A9, THBS4, and DNr63 for predicting the clinical outcome of a subject having a HPV⁺ cancer.

In a more preferred embodiment, the present invention concerns the biomarkers S100A9, THBS4 and DNr63 for predicting the clinical outcome of a subject having a HPV⁺ cancer.

In an even more preferred embodiment, the present invention concerns the biomarkers S100A9 and THBS4 for predicting the clinical outcome of a subject having a HPV⁺ cancer.

Expression level of Biomarkers

In a preferred embodiment, the above-mentioned method is performed on cancer cells from a cancer sample from a patient.

The above-mentioned method can also necessitate the use of a normal sample, as a way of comparison to the sample, preferably to the cancer sample. The normal sample can be a sample from the same patient or from another patient. The normal sample can also be from another patient, preferably a normal or healthy patient, i.e. a patient who does not suffer from a cancer.

The above-mentioned method may also comprise a step of obtaining or providing a sample from said patient.

The detection and expression level of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63 may be determined by any method known by the skilled person in the art. In particular, detection may be done by highlighting the presence and/or the quantity of proteins, and/or expression level may be determined by measuring the quantity of mRNA.

In a particular embodiment, the expression level of the biomarker(s) S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and/or DNr63 is determined by measuring the quantity of the mRNA transcripts, for instance by quantitative RT-PCR, real time quantitative RT-PCR or RNA-seq.

Methods for determining the quantity of mRNA are well known in the art and include, but are not limited to, quantitative or semi-quantitative RT-PCR, real time quantitative or semi- quantitative RT-PCR, Nanostring technology, sequencing based approaches, for instance by high- throughput sequencing technology such as RNA-Seq or sequencing technologies using microfluidic systems, or transcriptome approaches.

The nucleic acid contained in the sample (e.g., cells or tissue prepared from the patient) may be first extracted according to standard methods, for example using lytic enzymes or chemical solutions or extracted by nucleic-acid-binding resins following the manufacturer's instructions. These nucleic acids may be frozen to be stored before use.

The extracted mRNA may be then detected by hybridization (e.g., Northern blot analysis) and/or amplification (e.g., RT-PCR). Quantitative or semi-quantitative RT-PCR is preferred. Real time quantitative or semi-quantitative RT-PCR is particularly advantageous. Preferably, primer pairs were designed in order to overlap an intron, so as to distinguish cDNA amplification from putative genomic contamination. Such primers may be easily designed by the skilled person. Other methods of Amplification include, but are not limited to, ligase chain reaction (LCR), transcription-mediated amplification (TMA), strand displacement amplification (SDA) and nucleic acid sequence based amplification (NASBA).

Alternatively, the quantity of mRNA may also be measured using the Nanostring's NCOUNTER™ Digital Gene Expression System (Geiss et al. 2008 Nat. Biotechnol. 26:317-325) which captures and counts individual mRNA transcripts by a molecular bar-coding technology and is commercialized by Nanostring Technologies, or the QuantiGene^® Plex 2.0 Assay (Affymetrix). The quantity of mRNA may further be determined using approaches based on high- throughput sequencing technology such as RNA-Seq (Wang et al. Nat Rev Genet. 2009 January; 10(1): 57-63) or sequencing technologies using microfluidic systems.

The expression level of a gene may also be determined by measuring the quantity of mRNA by transcriptome approaches, in particular by using DNA microarrays. To determine the expression level of a gene, the sample, optionally first subjected to a reverse transcription, is labelled and contacted with the microarray in hybridization conditions, leading to the formation of complexes between target nucleic acids that are complementary to probe sequences attached to the microarray surface. The labelled hybridized complexes are then detected and can be quantified or semi-quantified. Labelling may be achieved by various methods, e.g. by using radioactive or fluorescent labelling. Many variants of the microarray hybridization technology.

Next Generation Sequencing methods (NGS) may also be used.

In a particular embodiment, the quantity of mRNA is measured by quantitative RT-PCR.

As used herein, the terms "quantitative RT-PCR", "qRT-PCR", "Real time RT-PCR" and "quantitative Real time RT-PCR" are equivalent and can be used interchangeably. Any of a variety of published quantitative RT-PCR protocols can be used (and modified as needed) for use in the present method. Suitable quantitative RT-PCR procedures include but are not limited to those presented in U.S. Pat. No. 5,618,703 and in U.S. Patent Application No. 2005/0048542, which are hereby incorporated by reference.

In a preferred embodiment of the above mentioned method, the quantitative RT-PCR includes two main steps, the reverse transcription (RT) of RNA in cDNA and the quantitative PCR (Polymerase Chain Reaction) amplification of the cDNA. Quantitative RT-PCR can be performed by an uncoupled or by a coupled procedure. In an uncoupled quantitative RT-PCR, the reverse transcription is performed independently from the quantitative PCR amplification, in separate reactions. Whereas, in a coupled quantitative RT-PCR, the reverse transcription and the quantitative PCR amplification are performed in a single reaction tube using a common reaction mixture including both the reverse transcriptase and the DNA polymerase. The method of the invention encompasses all versions of quantitative RT-PCR.

Preferably, expression levels of genes S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and/or DNr63 are normalized to a reference expression level, preferably to the expression level of one or more housekeeping (or control or reference) genes.

As used herein, the term "housekeeping gene" refers to a gene involved in basic functions needed for maintenance of the cell. Housekeeping genes are transcribed at a relative constant level and are thus used to normalize expression levels of genes that vary across different samples. Examples of housekeeping genes include, but are not limited to, GAPDH (Gene ID NCBI 2597), ribosomal 18S gene (RNA18S5, Gene ID NCBI: 100008588), beta-glucuronidase, b-actin (ACTB), peptidylprolyl isomerase A (cyclophilin A, PPIA), tubulin, ubiquitin, RPLPO, HPRT1 and B2M genes.

In a particular embodiment, the expression level of each gene is determined by measuring the amount of mRNA by quantitative RT-PCR and is normalized with respect to that of a housekeeping gene, for example the peptidylprolyl isomerase A (cyclophilin A, PPIA), b-actin (ACTB), and Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) reference genes. Preferably, a gene of interest is normalized by the geometric mean of the expression of two reference genes, the Ribosomal Protein Lateral stalk subunit P0 (RPLPO) and the Ubiquitin B (UBB) genes.

Quantitative RT-PCR Primers that can be used to measure S100A7 expression level in a sample are for example: Forward: 5'- GCCTGCTGACGATGATGAAG -3' (SEQ ID NO: 1) and Reverse: 5’- ATGGCTCTGCTTGTGGTAGT -3' (SEQ ID NO: 2).

Quantitative RT-PCR Primers that can be used to measure S100A9 expression level are for example: Forward: 5'- CAAAATGTCGCAGCTGGAAC -3' (SEQ ID NO: 3) and Reverse: 5’- CATTT GT GTCCAGGT CCT CC -3' (SEQ ID NO: 4).

Quantitative RT-PCR Primers that can be used to measure SERPINB1 expression level in a sample are for example: Forward: 5'- AAGTTTGGCTCTGTTGGCTGT 3' (SEQ ID NO: 5) and Reverse: 5’- TCCCATGGCTATCAGGAGGA -3' (SEQ ID NO: 6).

Quantitative RT-PCR Primers that can be used to measure SPRR1A expression level in a sample are for example: Forward: 5'- AGTTAGCATGCTGTCACCCT -3' (SEQ ID NO: 7) and Reverse: 5’- CATCCTCAAATGCACCCGAG -3' (SEQ ID NO: 8).

Quantitative RT-PCR Primers that can be used to measure KRT6B expression level in a sample are for example: Forward: 5'-TCTAGGTCCAGCTGCAGATG-3' (SEQ ID NO: 9) and Reverse: 5'- GAGAGCAGAGAAAGCAGTGC-3' (SEQ ID NO: 10).

Quantitative RT-PCR Primers that can be used to measure THBS4 expression level in a sample are for example: Forward: 5'- GCAGACAGAGATGGCATTGG -3' (SEQ ID NO: 11) and Reverse: 5’- ATCG GTGT CTTT CTG GTCGT -3' (SEQ ID NO: 12).

Quantitative RT-PCR Primers that can be used to measure DNr63 expression level in a sample are for example: Forward: 5'- GAAGAAAGGACAGCAGCATTGA -3' (SEQ ID NO: 13) and Reverse: 5’- CTGGGCATTGTTTTCCAGGTA -3 (SEQ ID NO: 14).

In another particular embodiment, the S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and/or DNr63 proteins may be detected and/or measured by any method known by the skilled person in the art. Usually, these methods comprise contacting the sample with a binding partner capable of selectively interacting with the protein present in the sample. The binding partner is generally a polyclonal or monoclonal antibody, preferably a monoclonal antibody. Such an antibody can be produced through methods known to the man skilled in the art. This antibody includes in particular those produced by a hybridoma and those produced by genetic engineering using host cells transformed with a recombinant expression vector carrying a gene encoding the antibody. A hybridoma producing monoclonal antibodies can be obtained as following: the protein or immunogenic fragments thereof are used as antigens for immunization according to conventional methods of immunization. The resulting immunocytes are fused with known parent cells according to conventional cell fusion methods and the cells producing the antibodies are thus screened from fused cells using conventional screening methods.

The antibody according to the invention can be labelled and/or fused to a detection entity. Preferably, the antibody according to the invention is labelled or fused to a detection entity.

In a preferred embodiment, the antibody is labelled. The antibody can be labelled with a label selected from the group consisting in a radiolabel, an enzyme label, a fluorescent label, a biotin-avidin label, a chemiluminescent label, a gold label and the like. The antibody according to the invention can be labelled by standard labelling techniques well known by the man skilled in the art and labelled antibodies can be visualized using known methods. In particular, labels generally provide signals detectable by fluorescence, chemo-luminescence, radioactivity, colorimetry, mass spectrometry, X-ray diffraction or absorption, magnetism, enzymatic activity, or the like.

In another preferred embodiment, the above mentioned antibody can be fused to a detection entity. The detection entity may be selected from the group consisting of a tag, an enzyme or a fluorescent protein. Preferably, the detection entity is at the C-terminal extremity of the antibody.

In a particular embodiment, the biomarker(s) S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and/or DNr63 is/are detected by identifying the presence and/or the quantity of proteins, for instance by semi-quantitative Western blots, enzyme-labelled and mediated immunoassays, such as ELISAs, biotin/avidin type assays, radioimmunoassay, immuno-electrophoresis or immunoprecipitation or by protein or antibody arrays, mass spectrometry and flow cytometry (FACS) or by immunohistochemistry, in particular by using antibodies specific of human S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and/or DNr63.

Immunohistochemistry (IHC) refers to the process of selectively imaging antigens (e.g. proteins) in cells of a tissue section by exploiting the principle of antibodies binding specifically to antigens in biological tissues. Visualizing the antibody-antigen interaction can be accomplished in a number of ways, well known by the man skilled in the art. In the most common instance, an antibody is conjugated to an enzyme, such as peroxidase, that can catalyse a colour-producing reaction or is tagged by a fluorophore, such as fluorescein or rhodamine. Immunohistochemistry can be divided into two phases: sample preparation and sample labelling.

Preparation of the sample is critical to maintain cell morphology, tissue architecture and the antigenicity of target epitopes. This requires proper tissue collection, fixation and sectioning. A solution of paraformaldehyde is often used to fix tissue, but other methods may be used. The tissue may then be sliced or used whole, dependent upon the purpose of the experiment or the tissue itself. Before sectioning, the tissue sample may be embedded in a medium, like paraffin wax or cryomedia. Sections can be sliced on a variety of instruments, most commonly a microtome, cryostat, or Compresstome tissue slicer. Specimens are typically sliced at a range of 3 miti-50 pm. The slices are then mounted on slides, dehydrated using alcohol washes of increasing concentrations (e.g., 50%, 75%, 90%, 95%, 100%), and cleared using a detergent like xylene before being imaged under a microscope. Depending on the method of fixation and tissue preservation, the sample may require additional steps to make the epitopes available for antibody binding, including deparaffinization and antigen retrieval.

For formalin-fixed paraffin-embedded tissues, antigen-retrieval is often necessary, and involves pre-treating the sections with heat or protease. These steps may make the difference between the target antigens staining or not staining. Depending on the tissue type and the method of antigen detection, endogenous biotin or enzymes may need to be blocked or quenched, respectively, prior to antibody staining. Although antibodies show preferential avidity for specific epitopes, they may partially or weakly bind to sites on nonspecific proteins (also called reactive sites) that are similar to the cognate binding sites on the target antigen. To reduce background staining in IHC, samples are incubated with a buffer that blocks the reactive sites to which the primary or secondary antibodies may otherwise bind. Common blocking buffers include normal serum, non-fat dry milk, BSA, or gelatine. Methods to eliminate background staining include dilution of the primary or secondary antibodies, changing the time or temperature of incubation, and using a different detection system or different primary antibody. Quality control should as a minimum include a tissue known to express the antigen as a positive control, and negative controls of tissue known not to express the antigen, as well as the test tissue probed in the same way with omission of the primary antibody (or better, absorption of the primary antibody).

For immunohistochemical detection strategies, antibodies are classified, when necessary, as primary or secondary reagents. Primary antibodies are raised against an antigen of interest and are typically unconjugated (i.e. unlabelled), while secondary antibodies are raised against immunoglobulins of the primary antibody species. The secondary antibody is usually labelled and/or fused to a detection entity as described above.

The direct method is a one-step staining method and involves a labelled antibody reacting directly with the antigen in tissue sections. While this technique utilizes only one antibody and therefore is simple and rapid, the sensitivity is lower due to little signal amplification, in contrast to indirect approaches.

The indirect method involves an unlabelled primary antibody (first layer) that binds to the target antigen in the tissue and a labelled secondary antibody (second layer) that reacts with the primary antibody. The secondary antibody must be raised against the IgG of the animal species in which the primary antibody has been raised. This method is more sensitive than direct detection strategies because of signal amplification due to the binding of several secondary antibodies to each primary antibody if the secondary antibody is conjugated to the fluorescent or enzyme reporter. Further amplification can be achieved if the secondary antibody is conjugated to several biotin molecules, which can recruit complexes of avidin-, streptavidin- or NeutrAvidin protein- bound enzyme. The antibodies can be for example, the antibody anti-S100A7 clone HPA006997 from Sigma, the antibody anti-S100A9 clone HPA004193 from Sigma, the antibody anti-KRT6B clone PA5-29134 from ThermoFisher Scientific, the antibody anti-THBS4 clone G10 from Santa Cruz- 28293.

The level of biomarker in the cancer sample is compared to a reference expression level.

In particular, the reference expression level can be the expression levels of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B or DNr63 in a normal sample. The normal sample is a non tumour sample. Expression levels obtained from cancer samples and normal samples may be normalized by using expression levels of proteins which are known to have stable expression, e.g. the housekeeping gene as detailed above. Alternatively, the reference expression level may be the expression level of a gene having a stable expression in cancer samples. Such genes include for example, RPLPO, TBP, GAPDFI, UBB or a-actin. Preferably, the reference expression level is the geometric mean of the expression levels of the RPLPO and UBB genes.

The method comprises the step of determining whether the expression levels of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B or DNr63 are dysregulated compared to the reference expression level, e.g. up-regulated or down regulated.

The expression levels of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B or DNr63 in the cancer sample are considered as significantly different (i.e. dysregulated, e.g. up or down regulated) compared to the reference expression levels in a normal sample, if, after normalization, differences are in the order of at least 2-fold (or more) higher or lower for the considered biomarker than the expression levels in the normal sample. Preferably, the expression levels of the biomarker(s) S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B or DNr63 in the cancer sample are considered as dysregulated if the levels are at least 3-fold higher or lower, or 4, 5 or 6-fold higher for the cancer sample than the expression levels in the normal sample.

Accordingly "high" means that the biomarker's level is at least 2, 3, 5 or 6-fold higher in the cancer sample than the expression levels in the normal sample. On the opposite, "low" means that the biomarker's level is at least 2, 3, 5 or 6-fold lower in the cancer sample than the expression levels in the normal sample.

If the reference expression level is the expression level of a gene having a stable expression in different cancer samples, in particular the expression level of the RPLPO and UBB gene, the expression level of the biomarker(s) is considered as high if the level or quantity of biomarker(s) mRNA is of at least or about 0.2, 0.25, 0.3, 0.35, 0.4, 0.45, 0.5, 0.55, 0.6, 0.65, 0.7, 0.75, 0.8, 0.85, 0.9, 0.95 or 1% of the RPLPO mRNA level or quantity reference or in the range of 0.2-1% of the RPLPO mRNA level or quantity reference. The level or quantity of the biomarker(s) S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and/or DNr63 relative to the quantity of RPLPO and UBB mRNA may be calculated by applying x=Cp Gene /Geometric mean (Cp RPLPO Cp UBB), where "Cp" means Crossing point. The exact value should be determined using a reference kit. Specifically, the biomarker(s) S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B or DNr63 mRNA levels should be determined in a number of reference cell lines with the same primers, reference gene and QPCR amplification kit that are intended to be used with the tumour samples. A bivariate analysis can be used to compare gene expression levels in metastatic vs non metastatic tumours: the minimal, maximal and median gene expression levels can be compared and a two sample Wilcoxon rank- sum (Mann-Whitney) test can be used to determine if the minimal, maximal and median gene expression levels are significantly different in the two categories of tumours. The genes that are statistically associated to distant metastasis in this analysis can further be studied with a ROC curve analysis, and a cut-off value may then be determined using the Liu method (which defines the optimal cut-point as the point maximizing the product of sensitivity and specificity). This cut-off value may dependent on expression levels of the biomarker(s) in the various cell types containing both normal and tumour cells. This cut-off value may be easily adjusted by the skilled person using another reference gene. This value may vary depending on the type of cancer and will be easily adapted by the one skilled in the art. The cut-off value will be chosen so as to obtain a significant p-value.

Typically, the accuracy of the test to discriminate diseased cases from normal cases, or a stage of cognitive impairment from another one, may be evaluated using Receiver Operating Characteristic (ROC) curve analysis (Metz CE, Semin Nucl Med. 1978 Oct; 8(4):283-98; Zweig & Campbell, Clin Chem. 1993 Apr; 39(4):561-77). In signal detection theory, a ROC curve, is a graphical plot of the sensitivity (or true positive rate), versus false positive rate (1 - specificity or 1 - true negative rate), for a binary classifier system. Each point on the ROC plot represents a sensitivity/specificity pair corresponding to a particular decision threshold. The area under the ROC curve (AUC) is a measure of how well a parameter can distinguish between two diagnostic groups (for example "diseased"/"normal" or "good prognosis/"bad prognosis" or "susceptible to develop metastasis"/"not susceptible to develop metastasis"). In a particular embodiment, the area under the ROC curve (AUC) is a measure of how well a parameter can distinguish between metastatic patients and non-metastatic patients.

For example, the reference value can be expressed as a concentration of the biomarker in the biological sample of the tested subject for a particular specificity and/or sensitivity, or can be a normalized cut-off value expressed as a ratio for a particular specificity and/or sensitivity. If a higher or lower sensitivity and/or specificity is/are desired, the cut-off value can easily be changed by the skilled person in the art, for example using a different reagent for a particular biomarker.

Preferably, the person skilled in the art may compare the expression level (obtained according to the method of the invention) with a defined threshold value. In one embodiment of the present invention, the threshold value is derived from the expression level (or ratio, or score) determined in a biological sample derived from one or more patients having HPV positive cancer. Furthermore, retrospective measurement of the expression level (or ratio, or scores) in properly banked historical patient samples may be used in establishing these threshold values.

To assess the performance of the models, the AUC (Area Under the Curve) of the receiver operating characteristics curves (ROC) were computed. The probability of overall survival was further estimated using the Kaplan Meier method. A log rank test was performed to test the difference between high- and low-risk groups. High and low-risk prognostic groups were defined according to the cut-off of returned probabilities of 0.5 : high risk having a returned probability > 0.5 of death within the first three years and low risk having a returned probability < 0.5 of death within the first three years.

In one embodiment, the expression of the biomarker selected in the group consisting of TFIBS4, S100A9, S100A7, SPRR1A and SERPINB1 or any combination thereof is determined by qPCR.

In another embodiment, the expression of the biomarker selected in the group consisting of S100A9, S100A7 and DNr63 or any combination thereof is determined by I HC. Cancer and Metastasis

The method of the invention is aimed to predict the clinical outcome of a subject having a HPV positive cancer.

In one embodiment, the HPV positive tumour is from a cancer selected from the group consisting of head and neck cancers, oropharyngeal cancers, hypopharynx cancers, oesophageal cancers, cervical cancer, anal cancer, vaginal cancers, vulvar cancers, penile cancers, endometrial cancers and uterine cancers.

Preferably, the HPV positive cancer is selected from the group consisting of head and neck cancers, oropharyngeal cancers, hypopharynx cancers, oesophageal cancers. More preferably, the oropharyngeal cancers are throat, tongue or tonsils cancers.

Even more preferably, the HPV positive cancer is an oropharyngeal squamous cell carcinoma (OSCC).

In one embodiment the subject is suffering from cancer metastasis, preferably distant metastasis. In another embodiment, the tumour is identified as coming from Primary tumour (T),

Regional lymph nodes (N) or distant metastasis (M). The stage I to IV is known to the skilled person in the art to refer to the size and/or extent of the tumour. The higher the number after the T, N or M, the larger the tumour or the more it has grown into nearby tissues has spread to other parts of the body. In a particular embodiment, the above characteristics of the tumour can be combined to the biomarkers selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63 for patient prognosis determination.

In a preferred embodiment, the above characteristics of the tumour can be combined to DNr63 for patient prognosis determination. In another preferred embodiment, the above characteristics of the tumour can be combined to S100A9 and/or THBS4 for patient prognosis determination.

In another preferred embodiment, the above characteristics of the tumour can be combined to S100A9 and/or THBS4 and/or DNr63 for patient prognosis determination.

Patient The patient is an animal, preferably a mammal, even more preferably a human. However, the patient can also be a non-human animal, in particular mammals such as dogs, cats, horses, cows, pigs, sheep, donkeys, rabbits, ferrets, gerbils, hamsters, chinchillas, rats, mice, guinea pigs and non-human primates, among others, that are in need of treatment.

The human patient according to the invention may be a human at the prenatal stage, a new-born, a child, an infant, an adolescent or an adult, in particular an adult of at least 30 years old, preferably an adult of at least 40 years old, still more preferably an adult of at least 50 years old, even more preferably an adult of at least 60 years old.

Preferably, the patient has been diagnosed with a cancer.

In a particular embodiment, the patient has already received at least one line of treatment, by any method known by the skilled person in the art, e.g. surgery, chemo-therapy, radiotherapy or combination thereof.

In a particular embodiment, the gender, the age, the tobacco smoking history, the tumour stage and/or size of the patient are determined and/or are combined to the biomarkers selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63 for patient prognosis determination. Prognosis

In one embodiment, the invention concerns the use of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63 for predicting the clinical outcome of a subject having a HPV positive cancer.

In a particular embodiment, the invention concerns an in vitro method for predicting the clinical outcome of a subject having a HPV positive cancer. The method comprises the detection of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63 in a cancer sample from said subject, and/or the determination of the expression level of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63 in said cancer sample. The expression level of the biomarkers is indicative of the clinical outcome of the patient e.g. patient survival, cancer recurrence and/or metastasis occurrence, preferably distant metastasis occurrence.

Good prognosis is preferably a good survival prognosis, a good metastasis-free survival, a decrease disease recurrence and/or a decrease metastasis occurrence, preferably a decrease of distant metastasis occurrence.

Poor prognosis is preferably a poor survival prognosis, an early disease progression, a lymph node involvement, an increased disease recurrence, especially after resection and/or treatment, or more preferably an increased metastasis occurrence, even more preferably an increase of distant metastasis occurrence.

Preferably, the method comprises providing a cancer sample from a patient having HPV⁺ cancer. The cancer and the patient are as described above.

Preferably, the metastasis occurrence is established 3 years after cancer treatment by any method known by the person skilled in the art, preferably, surgery, radiotherapy, chemotherapy, targeted therapy, immunotherapy or any combinations thereof.

Preferably, the survival of the patient is established 5 years after cancer treatment and/or cancer remission.

In a first aspect, high level of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, KRT6B and/or DNr63 is indicative of a good prognosis for the subject.

In a second aspect, low level of THBS4 is indicative of a good prognosis for the subject.

In a third aspect, high level of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, KRT6B and/or DNr63 and low level of THBS4 are indicative of a good prognosis for the subject.

In a fourth aspect, low level of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, KRT6B and/or DNr63 is indicative of a poor prognosis for the subject, preferably low level of S100A9.

In a fifth aspect, high level of THBS4 is indicative of a poor prognosis for the subject.

In a sixth aspect, low level of S100A9 and high level of THBS4 are predictive of a poor diagnosis.

In a seventh aspect, low level of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, KRT6B and/or DNr63 and high level of THBS4 are indicative of a poor prognosis for the subject.

Subject susceptible to benefit from a de-escalation therapy

In a particular embodiment, the invention concerns a method for selecting patient susceptible to benefit from a therapeutic de-escalation. The method comprises the determination of the clinical outcome of a subject having a HPV⁺ cancer by detecting at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63 in a cancer sample from said subject, and/or determining the expression level of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63 in said cancer sample, the expression level being indicative of the clinical outcome; and selecting patient with a good prognosis as patient susceptible to benefit from a therapeutic de-escalation. The method may further comprise a step of administering a therapeutic de-escalation (e.g., a reduced doses or frequency or an alternative therapeutic agent).

In a particular embodiment, high level(s) of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, KRT6B and/or DNr63 and/or low level of THBS4 are predictive of a good prognosis. A good prognosis is preferably a good survival prognosis, a good metastasis-free survival, a decrease disease recurrence and/or a decrease metastasis occurrence, more preferably a decrease of distant metastasis occurrence.

The results of the cancer sample analysis following the above method allows to discriminate patients with good prognosis, e.g. with lower risk of cancer recurrence and metastasis. This patient can be referred to a therapeutic de-escalation for a better comfort and an improved lifespan.

In one aspect, the therapeutic de-escalation can be a decreased doses or frequency of a treatment to the patient, decreasing the risk of side effects, or simplifying the treatment procedures. In particular, the therapeutic de-escalation can be a decreased doses or frequency of chemotherapy, in particular with Platinum-based antineoplastic drugs such as cisplatin, carboplatin, oxaliplatin, or nedaplatin. For cisplatin, decreased doses are less than 100 mg/m², for instance less than 70, 60, 50 or 40 mg/m², in particular about 40 mg/m². In addition or in combination, the therapeutic de-escalation can be a decreased doses or frequency of radiotherapy (e.g., 90, 80, 70, 60, 50 % of a standard dose radiation).

In another aspect, the therapeutic de-escalation can be the replacement of the chemotherapy or the combination of chemotherapy and radiotherapy by an alternative therapeutic agent which is less toxic. The alternative therapeutic agent can be used in combination with a radiotherapy or without radiotherapy. The alternative therapeutic agent can be used in combination with a chemotherapy or without, in particular with Platinum-based antineoplastic drugs such as cisplatin, carboplatin, oxaliplatin, or nedaplatin. The radiotherapy can be used at standard dose radiation (e.g., 50-70 Gy) or decreased dose radiation (e.g., 90, 80, 70, 60, 50 % of a standard dose radiation). The alternative therapeutic agent can be an epidermal growth factor receptor (EGFR) inhibitor such as cetuximab, panitumumab, gefitinib, erlotinib, afatinib, lapatinib, brigatinib icotinib and mereletinib. In one embodiment, the therapeutic de-escalation is the replacement of the chemotherapy or the combination of chemotherapy and radiotherapy by cetuximab. As an illustration, it can be referred to a review about the de-escalation in HPV-positive oropharyngeal carcinoma (Mirghani et al, 2015, Int J Cancer, 136, 1494-1503), the disclosure of which being incorporated herein by reference.

Accordingly, the present invention relates to a method for treating a cancer in a patient comprising selecting the patient susceptible to benefit from a therapeutic de-escalation by the method as disclosed herein and administering a therapeutic de-escalation. Preferably, the therapeutic de-escalation comprises the administration of a therapeutic efficient amount of an epidermal growth factor receptor (EGFR) inhibitor, in particular selected from the group consisting of cetuximab, panitumumab, gefitinib, erlotinib, afatinib, lapatinib, brigatinib icotinib and mereletinib, more preferably cetuximab. The epidermal growth factor receptor (EGFR) inhibitor can used in combination with a radiotherapy and/or a chemotherapy, especially with Platinum-based antineoplastic drugs such as cisplatin, carboplatin, oxaliplatin, or nedaplatin, more preferably cisplatin. More preferably, the radiotherapy and/or chemotherapy are used with a decreased regimen (decreased doses and/or frequencies).

Subject susceptible to benefit from an immunotherapy

In a particular embodiment, the invention concerns a method for selecting patient susceptible to benefit from an immunotherapy or susceptible to better respond to an immunotherapy.

The method comprises the determination of the clinical outcome of a subject having a FIPV⁺ cancer by detecting at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, TFIBS4, KRT6B and DNr63 in a cancer sample from said subject, and/or determining the expression level of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63 in said cancer sample, the expression level being indicative of the clinical outcome; and selecting patient with a good prognosis as patient susceptible to better respond to an immunotherapy.

The method may also comprise determining the expression level of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, TFIBS4, KRT6B and DNr63 in said cancer sample, the expression level being indicative of the response of the patient to an immunotherapy; and selecting patient susceptible to better respond to an immunotherapy.

In a particular embodiment, high level(s) of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, KRT6B and/or DNr63 and/or low level of THBS4 are predictive of a good prognosis. A good prognosis is preferably a good survival prognosis, a good metastasis-free survival, a decrease disease recurrence and/or a decrease metastasis occurrence, more preferably a decrease of distant metastasis occurrence. The biomarkers can be used in combination, in particular according to any combination disclosed herein.

The results of the cancer sample analysis following the above method allows to discriminate patients with good prognosis, e.g. with lower risk of cancer recurrence and metastasis and also patient likely to respond better to immunotherapy.

Such immunotherapy can be based on monoclonal antibodies, including checkpoint inhibitors, bispecific antibodies, adoptive cell transfer and anti-tumour vaccination. In particular, the immunotherapeutic compound can be selected in the group consisted of Nivolumab, Ipilimumab, Pembrolizumab, Durvalumab, Avelumab, Urelumab, Tremelimumab, therapeutic vaccination and combination thereof.

Kit and Uses thereof

In another aspect, the invention concerns a kit and its uses, the kit comprising means for measuring the expression level of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63. The kit can be used for (i) predicting the clinical outcome of a subject having a HPV positive cancer (ii) selecting a subject affected with HPV positive cancer with good or poor prognosis and/or (iv) determining whether a subject affected with a HPV positive cancer is susceptible to benefit from therapeutic de-escalation or to better respond to immunotherapy.

For instance, the means suitable for determining the expression levels of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and/or DNr63 can be antibodies, primers and/or probe specific to S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63 proteins and/or genes; and, optionally, a leaflet providing guidelines to use such a kit.

In a first aspect, the kit comprises (i) at least one antibody specific to at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, and/or DNr63 and, optionally, means for detecting the formation of the complexes between S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and/or DNr63; and/or

(ii) at least one probe specific to mRNA or cDNA of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and/or DNr63, and, optionally, means for detecting the hybridization of said at least one probe on S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and/or DNr63 mRNA or cDNA; and/or (iii) at least one nucleic acid primer pair specific to at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and/or DNr63 mRNA or cDNA and, optionally, means for amplifying and/or detecting said mRNA or cDNA ; and,

(iv) optionally, a leaflet providing guidelines to use such a kit.

In a preferred embodiment, the kit comprises:

(i) at least one antibody specific to each of the followings: S100A9 and THBS4 and, optionally, means for detecting the formation of the complexes between S100A9 and THBS4; and/or

(ii) at least one antibody specific to each of the followings : S100A9 and THBS4 and, optionally, means for detecting the formation of the complexes between S100A9 and THBS4; and/or

(iii) at least one probe specific to mRNA or cDNA of S100A9 and THBS4, and, optionally, means for detecting the hybridization of said at least one probe on S100A9 and THBS4 mRNA or cDNA; and/or

(iv) at least one nucleic acid primer pair specific to S100A9 and THBS4 mRNA or cDNA and, optionally, means for amplifying and/or detecting said mRNA or cDNA ; and,

(v) optionally, a leaflet providing guidelines to use such a kit.

In another preferred embodiment, the kit comprises:

(i) at least one antibody specific to each of the followings: S100A9, THBS4 and DNr63 mRNA and, optionally, means for detecting the formation of the complexes between S100A9, THBS4 and DNr63; and/or

(ii) at least one antibody specific to each of the followings : S100A9, THBS4 and DNr63 and, optionally, means for detecting the formation of the complexes between S100A9, THBS4 and DNr63; and/or

(iii) at least one probe specific to mRNA or cDNA of S100A9, THBS4 and DNr63, and, optionally, means for detecting the hybridization of said at least one probe on S100A9, THBS4 and DNr63 mRNA or cDNA; and/or

(iv) at least one nucleic acid primer pair specific to S100A9, THBS4 and DNr63 mRNA or cDNA and, optionally, means for amplifying and/or detecting said mRNA or cDNA ; and,

(v) optionally, a leaflet providing guidelines to use such a kit. In another preferred embodiment, the kit comprises (i) at least one antibody specific to S100A9 and, optionally, means for detecting the formation of the complexes between S100A9; and/or (ii) at least one antibody specific to S100A9 and, optionally, means for detecting the formation of the complexes between S100A9; and/or (iii) at least one probe specific to mRNA or cDNA of S100A9, and, optionally, means for detecting the hybridization of said at least one probe on S100A9 mRNA or cDNA; and/or (iv) at least one nucleic acid primer pair specific to S100A9 mRNA or cDNA and, optionally, means for amplifying and/or detecting said mRNA or cDNA ; and, (v) optionally, a leaflet providing guidelines to use such a kit.

In another preferred embodiment, the kit comprises (i) at least one antibody specific to S100A7 and, optionally, means for detecting the formation of the complexes between S100A7; and/or (ii) at least one antibody specific to S100A7 and, optionally, means for detecting the formation of the complexes between S100A7; and/or (iii) at least one probe specific to mRNA or cDNA of S100A7, and, optionally, means for detecting the hybridization of said at least one probe on S100A7 mRNA or cDNA; and/or (iv) at least one nucleic acid primer pair specific to S100A7 mRNA or cDNA and, optionally, means for amplifying and/or detecting said mRNA or cDNA ; and, (v) optionally, a leaflet providing guidelines to use such a kit.

In another preferred embodiment, the kit comprises (i) at least one antibody specific to SERPINB1 and, optionally, means for detecting the formation of the complexes between SERPINB1; and/or (ii) at least one antibody specific to SERPINBland, optionally, means for detecting the formation of the complexes between SERPINB1; and/or (iii) at least one probe specific to mRNA or cDNA of SERPINB1, and, optionally, means for detecting the hybridization of said at least one probe on SERPINBlmRNA or cDNA; and/or (iv) at least one nucleic acid primer pair specific to SERPINBlmRNA or cDNA and, optionally, means for amplifying and/or detecting said mRNA or cDNA ; and, (v) optionally, a leaflet providing guidelines to use such a kit.

In another preferred embodiment, the kit comprises (i) at least one antibody specific to SPRR1A and, optionally, means for detecting the formation of the complexes between SPRR1A; and/or (ii) at least one antibody specific to SPRR1A and, optionally, means for detecting the formation of the complexes between SPRR1A; and/or (iii) at least one probe specific to mRNA or cDNA of SPRR1A, and, optionally, means for detecting the hybridization of said at least one probe on SPRR1A mRNA or cDNA; and/or (iv) at least one nucleic acid primer pair specific to SPRR1A mRNA or cDNA and, optionally, means for amplifying and/or detecting said mRNA or cDNA ; and, (v) optionally, a leaflet providing guidelines to use such a kit.

In another preferred embodiment, the kit comprises (i) at least one antibody specific to THBS4 and, optionally, means for detecting the formation of the complexes between THBS4; and/or (ii) at least one antibody specific to THBS4 and, optionally, means for detecting the formation of the complexes between THBS4; and/or (iii) at least one probe specific to mRNA or cDNA of THBS4, and, optionally, means for detecting the hybridization of said at least one probe on THBS4 mRNA or cDNA; and/or (iv) at least one nucleic acid primer pair specific to THBS4 mRNA or cDNA and, optionally, means for amplifying and/or detecting said mRNA or cDNA ; and, (v) optionally, a leaflet providing guidelines to use such a kit.

In another preferred embodiment, the kit comprises (i) at least one antibody specific to KRT6B and, optionally, means for detecting the formation of the complexes between KRT6B; and/or (ii) at least one antibody specific to KRT6B and, optionally, means for detecting the formation of the complexes between KRT6B; and/or (iii) at least one probe specific to mRNA or cDNA of KRT6B, and, optionally, means for detecting the hybridization of said at least one probe on KRT6B mRNA or cDNA; and/or (iv) at least one nucleic acid primer pair specific to KRT6B mRNA or cDNA and, optionally, means for amplifying and/or detecting said mRNA or cDNA ; and, (v) optionally, a leaflet providing guidelines to use such a kit.

In another preferred embodiment, the kit comprises (i) at least one antibody specific to DNr63 and, optionally, means for detecting the formation of the complexes between DNr63; and/or (ii) at least one antibody specific to DNr63 and, optionally, means for detecting the formation of the complexes between DNr63; and/or (iii) at least one probe specific to mRNA or cDNA of DNr63, and, optionally, means for detecting the hybridization of said at least one probe on DNr63 mRNA or cDNA; and/or (iv) at least one nucleic acid primer pair specific to DNr63 mRNA or cDNA and, optionally, means for amplifying and/or detecting said mRNA or cDNA; and, (v) optionally, a leaflet providing guidelines to use such a kit.

All the references cited in this application, including scientific articles and summaries, published patent applications, granted patents or any other reference, are entirely incorporated herein by reference, which includes all the results, tables, figures and texts of theses references.

Although having different meanings, the terms "comprising", "having", "consisting in" and "containing" can be replaced one for the other in the entire application.

Further aspects and advantages of the present invention will be described in the following examples, which should be regarded as illustrative and not limiting. EXAMPLES

RESULTS

Comparative analysis of HPV positive cancer gene expression

In order to characterize HPV-positive cancer, the inventors performed an analysis of transcriptomic data from 8 tumours samples. The principal component analysis of the data uncovered two different clusters of tumours corresponding to 6 and 2 patients, respectively (Fig. 1A). A hierarchical clustering analysis of gene expression data identified 128 genes to be significantly differentially expressed between cluster 1 and cluster 2 (Fig. IB). Genes that are upregulated and downregulated in cluster 1 compared to cluster 2 with a log-fold-change > 1.5 (adjusted p-value <5.0E-03) are shown in table 1 and table 2, respectively.

Meta-analysis of publically available transcriptomic data sets previously generated were also analysed and recovered cluster 1 and cluster 2 in transcriptomic data from 4 independent studies, with correlation coefficients ranking from 0.21 to 0.64, and 0.42 to 0.55, respectively (Table 3). In order to assess if the molecular identity of cluster 1 and cluster 2 correlated with different prognosis, a Kaplan-Meier analysis of the overall survival of the 8 FlPV-positive patients was performed. Two patients from cluster 2 displayed a shorter survival (log-rank test p-value=0.004; Fig. 1C).

The inventors identified several genes upregulated in cluster 1 compared to cluster 2. These genes included the biomarkers S100A7, S100A9, KRT6A, KRT6B, SPRR3, SPRR1A, SPRR1B, SERPIN Bl, DEFB4A and CLCA2 (Table 1). Interestingly, these genes are regulated by the DNr63 transcription factor. The inventor measured the correlation between the expression of the genes described and the expression of TP63 (which encodes the DNr63 transcription factor).

In order to confirm this hypothesis, the expression of the DNr63 and of the above biomarkers was analysed, both at the gene and protein level, in FlPV-positive oral squamous cell carcinoma samples (OSCC). A cohort of 77 patients was constituted by selecting tumour specimen from the tumour banks of the Paul Strauss Cancer Center (N=34) and of the Liege University Hospital (N=43). The detailed demographics and pathological features of both cohorts, as well as their combination are shown in Table 4. The two patient series did not show any significant differences in terms of demographics, pathological features, treatment and outcome (Table 8).

Formalin-fixed paraffin-embedded tumour specimens from 71 patients were retrieved and slides were stained with commercial antibodies raised against p63, S100A7, S100A9, KRT6B and TFIBS4. A semi-quantitative analysis of the staining was monitored according to signal intensity (no signal: 0; weak signal: 1; moderate signal: 2; strong signal: 3) and to the proportion of positively- labelled carcinoma cells (<5%: 0; 5%-33%: 1; 34-66%: 2; >67%: 3). The product of these two features allowed therefore ranking tumours along a scale from 0 to 9.

DNr63 is known to be the major p63 isoform expressed in normal oral epithelium as well as Head and Neck squamous-cell carcinoma (HNSCC). A positive signal in the nuclei of epithelial basal cells is observed in tumour associated normal epithelium. DNr63 staining was found to be heterogeneous among HPV-positive OSCC. Based on nuclear DNr63 expression, the tissue samples were classified into two groups: ANp63-low/neg (score ranged between 0 and 3; N=29/71) versus ANp63-high (score >4; N=42/71). The expression of S100A7, S100A9, KRT6B and THBS4 was measured in serial slides from the same specimens. Carcinoma cells in ANp63-high OSCC displayed a higher expression of S100A7, S100A9 and KRT6B, whereas a higher expression of THBS4 was observed in ANp63-low tumours (Fig. 3). A semi-quantitative analysis of the staining was carried- out, and the distribution of the IHC scores in the ANp63-low and -high tumours was performed (Table 9). High IHC score for S100A7, S100A9, KRT6B and low IHC scores for THBS4 were more frequently observed in ANp63-high HPV-positive OSCC (chi-2 p<0,001).

In addition, the expression of the S100A7, S100A9, KRTB6 and THBS4 genes was analysed using a RT-qPCR approach on RNA harvested from the same tumour samples, and the results were analysed by comparing ANp63-low and -high tumours. Similarly to what was observed at the protein level, ANp63-high OSCC displayed a lower expression of THBS4, and a higher expression of S100A7, S100A9 and KRTB6. These differences were statistically significant (Fig. 4).

The Biomarkers signature defines tumour subgroups of distinct prognosis

The IHC scores resulting from the semi-quantitative analysis of the S100A9, S100A7, KRT6B and THBS4 staining were analysed with respect to the occurrence of distant metastases during a 3- years follow-up period: the median IHC score, and minimal and maximal IHC score in metastatic vs. non-metastatic tumours was determined for each marker, and the two populations were compared. The S100A7 and S100A9 IHC score were significantly lower in metastasis-prone OSCC (Table 10). A ROC curve analysis was carried out and an optimal IHC score cut-off value that predicts distant metastasis was determined for S100A9, S100A7 and THBS4 (sensitivity specificity and AUC associated to each cut-off value are shown in Table 11). These cut-off values were included in a logistic regression model in order to measure the prognostic impact of the expression of the markers. High tumour expression of the THBS4 protein correlated with adverse prognosis, although the results did not reach statistical significance (Table 5). On the contrary, higher expression of S100A7 or S100A9 was found to correlate with a lower risk of metastatic spread (Table 5). A biomarker can be differentially expressed according to the target of interest (RNA or protein) or to the detection technique. For example, the expression of the KRT6B encoded gene measured by qPCR is associated with metastatic progression, whereas the expression of the KRT6B protein measured by IHC is not (p = 0.41).

A Kaplan-Meier analysis of the 3-years metastasis-free survival (MFS) showed that this stratification is significantly correlated with prognosis: patients with FlPV-positive ANp63-high OSCC have a prolonged MFS compared to their ANp63-low counterparts (p=0,012; Figure 5A). Tobacco smoke abuse and lymph nodes involvement have been previously shown to negatively impact on patients' outcome. More specifically, previous/current smokers with high-grade metastatic lymph node (N2b; N2c; N3) have adverse prognosis. Interestingly, stratification by DNr63 protein expression was able to discriminate patients with good vs. poor prognosis among this population (p=0.0008; Figure 5B).

The prognostic impact of the expression level of the genes on metastatic progression was also analysed. Therefore, a RT-qPCR approach was carried out on tumour RNA extracts and the expression of the DNr63, S100A7, S100A9, KRT6B, SERPINB1, SPRR1A, SPRR1B and THBS4 transcripts was monitored. Similarly to the analysis of the I HC score, the median, minimum and maximum expression level of each transcript was compared in metastatic and non-metastatic OSCC. Non-metastatic tumours were found to express significantly higher levels of S100A9, SERPINB1 and SPRR1A, close-to-significant higher expression of S100A7 and significantly lower levels of TFIBS4 (Table 6).

A ROC curve analysis was performed in order to determine an optimal gene expression cut off value that predicts distant metastasis for KRT6B, S100A7, S100A9, SERPINB1, SPRR1A, TFIBS4 and DNr63 (sensitivity specificity and Area under the curve (AUC) associated to each cut-off value are shown in Table 12). Using determined cut-off values, the prognostic power of these genes, alone or in combination, were evaluated.

The combination of the expression of TFIBS4 and S100A9 was found to be the most powerful predictor of metastatic spread in our patient series: a Kaplan-Meier analysis of the 3-year MFS showed that TFIBS4-high/S100A9-low tumour samples identify a population of poor-prognosis (p<0.0001; Figure 5C). A univariate cox regression analysis established that these patients are at high risk of metastatic progression compared to other TFIBS4-high/S100A9-high, TFIBS4- low/S100A9-low and THBS4-low/S100A9-high (OR=21.28; 95% Cl=5.38- 84.20; p<0.0001; Table 7). This predictor retained its prognostic power when it was included in a multivariate Cox regression model with potential confounding factors (gender; age; tumour stage; tobacco smoking; Table7). Table 1

Gene Log Fold p-value Adjusted Described to be

Change p-value regulated by DNr63

(Barbieri and collaborators)

S100A7

4.164956508 1.05E-08 2.62E-05

yes

SPRR1A 3.330663181 2.50E-09 1.25E-05 yes

CLCA4 3.136262718 4.49E-06 1.50E-03 no

TMPRSSUE 2.640565534 1.03E-06 6.44E-04 no

CRNN 2.454917609 2.76E-05 4.18E-03 no

C10orf99 2.414329705 1.91E-05 3.50E-03 no

TMPRSSUD 2.398586976 2.18E-05 3.76E-03 no

CLCA2 2.278068412 1.07E-05 2.81E-03 yes

CNFN 2.260158798 2.62E-05 4.18E-03 no

GJB6 2.128856976 4.36E-06 1.50E-03 no

TGM3 2.087250058 2.19E-06 1.09E-03 no

TGM1 1.983112569 1.37E-05 3.21E-03 no

KRT6A 1.699750425 5.73E-06 1.79E-03 yes

S100A9

1.672219459 2.66E-05 4.18E-03

Table 2

Gene Log Fold Change p-value Adjusted p-value

PROM1 -2.234553807 3.37E-05 4.81E-03

THBS4 -2.125087842 9.50E-06 2.64E-03

LOC100507263 -1.730934244 2.44E-06 1.11E-03

GRP -1.52920949 1.48E-05 3.21E-03

Table 3

Slebos et al. Tomar et al. Mirghani et al. Pyeon et al. (Affymetrix) (Affymetrix) (NimbleGen) (Affymetrix)

Clusterl 0.42 1.61E- 0.64 0.00 0.21 7.9E-03 0.31 1.94E-

08 05

Table 4

HPV-positive OSCC HPV-positive OSCC HPV-positive OSCC Strasbourg (N=34) Liege (N=43) Total (N=77)

Age

Age<60 years 16 (47%) 19 (47%) 35 (45%) Age>60 years 18 (53%) 24 (53%) 42 (55%)

Pathological tumour size

staging (pT): 5 (15%) 5 (15%) 10 (13%)

T1 16 (47%) 27 (47%) 43 (56%)

T2 12 (35%) 6 (35%) 18 (23%)

T3 1 (3%) 5 (3%) 6 (8%)

T4

Tumour stage

Stage 1 (3%) 3 (3%) 4 (5%) Stage 1 (3%) 6 (3%) 7 (9%) Stage 11 (32%) 12 (32%) 23 (30%) Stage 21 (62%) 22 (62%) 43 (56%)

Metastasis at 3 years

Yes 4 (12%) 5 (12%) 9 (12%)

No 29 (88%) 38 (88%) 67 (88%)

Table 5

Odds Ratio SE p-value 95% Cl

DNr63 IHC

0.157 0.133 0.028 0.030-0.823

S100A7 0.129 0.110 0.016 0.025-0.681

IHC

THBS4 IHC 2.073 0.259 0.394 0.388-11.068 Table 6

Metastasis N Median gene Minimum gene Maximum gene p-value at 3 years expression expression expression

rs No 68 1.060 0.010 13.028 0.064

1 Yes 9 0.133 0.002 2.883

No 30 1.048 0.285 2.551 0.010

CQ

5 Yes 4 0.286 0.126 0.879

s

CQ No 30 1 0.005 6.983 0.789

Yes 4 1.163 0.001 5.881

§£

CQ

No 29 1 0.179 4.830 0.152

2 Yes 4 0.489 0.344 2.090

;§ Table 7

Univariate Cox Regression Survival analysis (3 years MFS)

THBS4-high/S100A9-low 21.28012 14.93333 5.378278- <0.0001

84.1986

Hazard ratio SE 95% Cl p-value

Gender 0.91 0.8545205 0.14-5.73 0.920

Tumour Stage 3.27 3.791212 0.34-31.68 0.306

Table 8

HPV-positive OSCC

Strasbourg vs. Liege c² p-value

Age p=0.983

Pathological tumour size staging (pT) p=0.089

Tumour stage p=0.304

Metastasis at 3 years p=0.954

Table 9

KRTB6 IHC Score ANp63-high ANp63-low/neg.

HPV-positive OSCC HPV-positive OSCC

Table 9 (following)

THBS4 IHC Score ANp63-high ANp63-low/neg.

HPV-positive OSCC N=42 HPV-positive OSCC N=29

Table 10

DNr63 IHC _ No distant metastasis at 3 years Distant metastasis at 3 years

ANp63-low/neg. (N=29) N=22/29 (76%) N=7/29 (24%)

ANp63-high (N=42) _ N=40/42 (95%) _ N=2/42 (5%) x² p=0.027

Metastasis at 3 N Median S100A7 Min. S100A7 Max. S100A7

years IHC score IHC score IHC score

Yes 9 1 0 9

Mann-Whitney p=0.05

Metastasis at 3 N Median S100A9 Min. S100A9 Max. S100A9

years IHC score IHC score IHC score

Yes _ 9 _ 3 _ 1 _ 9

Mann-Whitney p=0.0027

Metastasis at 3 N Median THBS4 Min. THBS4 Max. THBS4

years IHC score IHC score IHC score

Yes _ 9 _ 6 _ 0 _ 6

Mann-Whitney p=0.252

Metastasis at 3 N Median KRT6B Min. KRT6B Max. KRT6B

years IHC score IHC score IHC score

Yes 9 6 3 6

Mann-Whitney p=0.419 Table 11

Empirical optimal Sensitivity Specificity AUC cutpoint

S100A7 IHC

2.5

0.69

0.78

0.73 S100A9 IHC 3.5 0.77 0.89 0.83 THBS4 IHC

5.5

0.67

0.69

0.68

Table 12

Empirical optimal Sensitivity Specificity AUC cutpoint

Material and methods

Gene expression assays:

List of primers

The qPCR conditions are as follows:

1. Pre-incubation : 5 min at 95°C

2. Amplification : 45 cycles • Denaturation: 10 seconds at 95°C

• Annealing: 10 seconds at 60°C

• Elongation : 10 seconds at 72°C

Gene expression was evaluated with respect to a standard curve.

Relative expression was normalized to the expression of two reference genes (Ubiquitin B (UBB) and Ribosomal Protein LP0 (RPLP0),· see above for sequence of primers and hybridization temperature (Tm)).

Immunohistochemistry: Paraffin-embedded tissue specimens of HPV-positive OSCC were collected. All samples were stained with anti-DeltaNp63 (anti-p40; Calbiochem, Gibbstown, NJ, USA), anti-S100A7 (clone HPA006997, Sigma), anti-S100A9 (clone HPA004193, Sigma), anti-Krt6 (clone PA5-29134, ThermoFisher Scientific) and anti-THBS4 (clone G10 Santa Cruz-28293) primary antibodies. Horseradish peroxidase (HRP)-conjugated secondary antibodies (either HRP-labelled polymer anti-mouse or HRP-labelled polymer anti-rabbit; Dako) were used to detect primary antibodies, and a colorimetric detection of HRP enzymatic activity was performed with the Dako EnVision kit. For each immunolabelled tissue, a semi-quantitative score of the intensity (0: negative; 1: low; 2: moderate; 3: intense) and extent (0: <5% of positive cells; 1: 6-33%; 2: 34-66%; 3: >67%) of the staining according to an arbitrary scale was used. As previously described (Hubert et al. J Pathol 2014), the results obtained with the two scales were multiplied and a global score (between 0 and 9) was obtained.

Statistical analyses:

• Unsupervised analysis: a descriptive Principal Component Analysis (R package factomineR, VI.32) was performed on transcriptomic data, normalized with the RMA (Robust Multi- Array Average) method and log2 transformed. An ascending hierarchical clustering approach allowing to uncover different groups of individuals was performed by using the Ward clustering method and a distance-based correlation

• Differential gene expression analysis: a moderate t-test (R package limma V3.22.7) was performed to uncover differentially expressed genes. Differential analysis of the transcriptome was performed using a moderate t-test.The family-wise error rate was controlled with the Bonferroni correction method. Results are presented as a heat map that highlights the relationship between the classification of samples and differentially expressed genes (function heat map, R package stats V3.1.2).

• Meta-analysis: publically available transcriptomic data sets were retrieved from the GEO (Gene Expression Omnibus) database. An unsupervised clustering analysis was performed on each retrieved data set, and centroids (corresponding to the median expression of each gene that was found to be differentially expressed in a given subgroup) were defined for clusters that were uncovered. Correlations between centroids were calculated (R package stats V3.1.2).

• Correlation matrix: the expression of genes described in the ANp63-dependent molecular signature described by Barbieri et al was analyzed and correlated to the expression of TP63 in our transcriptomic data set. The correlation matrix was established using the R corrplot function (package corrplot VO.77).

• The IHC scores were analyzed as follows: the median, maximum and minimum score for each marker was determined. The distribution of these features was compared in metastatic vs. non-metastatic tumors using a Pearson Chi2 and a Fisher's exact test. The cut-off value to discriminate metastatic and non-metastatic tumors was determined for each IHC score with the Liu Empirical cut point estimation method.

• The statistical relationship between the expression level of genes (median expression, minimum expression and maximal expression in metastatic vs. non-metastatic tumors) and the occurrence of distant metastasis was analyzed by using a Two-sample Wilcoxon rank- sum (Mann-Whitney) test. The cut-off value to discriminate metastatic and non-metastatic tumors was determined for each IHC score with the Liu Empirical cut point estimation method.

• The prognostic impact of protein and gene expression was analyzed by using: i) univariate Kaplan-Meier analysis (differences between survival curves were analyzed with a log-rank test; ii) multivariate Cox logistic regression.

Claims

1. Use of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, KRT6B and DNr63 for predicting the clinical outcome of a subject having a Human papillomavirus (HPV) positive cancer.

2. An in vitro method for predicting the clinical outcome of a subject having a HPV positive cancer, wherein the method comprises:

detecting at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4 and DNr63 in a cancer sample from said subject, and

3. The use according to claim 1 or the method according to claim 2, wherein the expression level of the biomarker is determined by measuring the quantity of the mRNA transcripts, for instance by quantitative RT-PCR, real time quantitative RT-PCR or RNA-seq.

4. The use according to claim 1 or 3 or method according to claims 2-3, wherein the expression level of the biomarker is determined by measuring the quantity of proteins, for instance by immunohistochemistry.

5. The use according to any one of claims 1 and 3-4 or method according to any one of claims 2-4, wherein the cancer is selected from the group consisting of head and neck cancers, oropharyngeal cancers, hypopharynx cancers, oesophageal cancers, cervical cancer, anal cancer, vaginal cancers, vulvar cancers, penile cancers, endometrial cancers and uterine cancers.

6. The use or method according to claim 5, wherein the oropharyngeal cancers are throat, tongue or tonsils cancers, preferably oropharyngeal squamous cell carcinoma.

7. The use according to any one of claims 1 and 3-6 or method according to claims 2- 6, wherein the expression level of the biomarker is combined with the gender and/or age of the subject, tobacco smoking, tumour stage and/or size for prognosis determination.

8. The use according to any one of claims 1 and 3-7 or method according to claims 2- 7, wherein high level of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, KRT6B and/or DNr63 and/or low level of THBS4 are predictive of a good prognosis, and wherein a good prognosis is preferably a good survival prognosis, a good metastasis-free survival, a decrease disease recurrence and/or a decrease metastasis occurrence.

9. The use or method according to claim 8, wherein a good prognosis is a decrease of distant metastasis occurrence.

10. The use according to any one of claims 1 and 3-9 or method according to claims 2- 9, wherein the biomarker is selected from the group consisting of S100A9, THBS4 and DNr63, alone or in combination.

11. The use according to any one of claims 1 and 3-10 or method according to claims 2-10, wherein the biomarkers are S100A9 and THBS4.

12. A method for selecting a subject susceptible to benefit from a therapeutic de- escalation, wherein the method comprises:

determining the clinical outcome of a subject having a Human papillomavirus positive cancer by the method according to any one of claims 2-11, and

selecting a subject with a good prognosis as susceptible benefit from a therapeutic de-escalation.

13. A method for selecting a subject susceptible to better respond to an immunotherapy, wherein the method comprises:

14. The method according to any one of claims 2-13,

wherein the biomarkers are S100A9 and THBS4,

wherein low level of S100A9 and high level of THBS4 are predictive of a poor diagnosis, and wherein a poor prognosis is a poor survival prognosis, an early disease progression, a lymph node involvement, an increased disease recurrence, especially after resection and/or treatment, or more preferably an increased metastasis occurrence.

15. Use of a kit comprising means for measuring the expression level of at least one biomarker selected from the group consisting of S100A9, S100A7, SERPINB1, SPRR1A, THBS4, and DNr63 for (i) predicting the clinical outcome of a subject having a HPV positive cancer (ii) selecting a subject affected with HPV positive cancer with a good or poor prognosis and/or (iv) determining whether a subject affected with a HPV positive cancer is susceptible to benefit from a therapeutic de-escalation or from an immunotherapy.