**Table 1.** Clinicopathological features of HNSCC patients.


analysis.

**Figure 1.** Analysis of classical markers for stratification of HNSCC samples. Statistical significance was indicated as \* for *p* < 0.05 and \*\* for *p* < 0.0001; (**A**) viral load quantification of HPV-positive HNSCC samples by qPCR; (**B**) differential p16 mRNA expression in HNSCC samples analyzed by qPCR; (**C**) Viral E7 mRNA expression in HPV-positive HNSCC samples; (**D**) Spearman correlation analyses between the expression of E7 (log2) oncogene and p16 (log2) showed correlation (r = 0.59; *p*  < 0.05) in HPV-positive HNSCC tumor samples. **Figure 1.** Analysis of classical markers for stratification of HNSCC samples. Statistical significance was indicated as \* for *p* < 0.05 and \*\* for *p* < 0.0001; (**A**) viral load quantification of HPV-positive HNSCC samples by qPCR; (**B**) differential p16 mRNA expression in HNSCC samples analyzed by qPCR; (**C**) Viral E7 mRNA expression in HPV-positive HNSCC samples; (**D**) Spearman correlation analyses between the expression of E7 (log<sup>2</sup> ) oncogene and p16 (log<sup>2</sup> ) showed correlation (r = 0.59; *p* < 0.05) in HPV-positive HNSCC tumor samples.

#### *3.2. p16 Gene Expression 3.2. p16 Gene Expression*

Forty-one out sixty-seven HNSCC matching DNA/RNA samples were available for further analyses. P16 mRNA expression was investigated in HNSCC samples, showing upregulation in 12/16 (75%) of HPV16-positive HNSCC and in 5/25 (20%) of HPV-negative HNSCC samples compared to controls (*p* < 0.001). HPV-positive patients showed overall p16 gene upregulation compared to controls (Mean ± [SD], 2.60 ± 3.98 log2 fold, *p* < 0.05) with the exception of two samples that harboured p16 downregulated. HPV-negative samples were downregulated compared to control samples (Mean ± [SD], −2.34 ± 3.71 log2 fold, *p* < 0.05). Differences in p16 expression between HPV-positive and -negative were also significant (*p* < 0.0001) (Figure 1B). Forty-one out sixty-seven HNSCC matching DNA/RNA samples were available for further analyses. P16 mRNA expression was investigated in HNSCC samples, showing upregulation in 12/16 (75%) of HPV16-positive HNSCC and in 5/25 (20%) of HPV-negative HNSCC samples compared to controls (*p* < 0.001). HPV-positive patients showed overall p16 gene upregulation compared to controls (Mean ± [SD], 2.60 ± 3.98 log<sup>2</sup> fold, *p* < 0.05) with the exception of two samples that harboured p16 downregulated. HPV-negative samples were downregulated compared to control samples (Mean ± [SD], −2.34 ± 3.71 log<sup>2</sup> fold, *p* < 0.05). Differences in p16 expression between HPV-positive and -negative were also significant (*p* < 0.0001) (Figure 1B).

#### HPV-positive HNSCC samples were analyzed for HPV16 E7 gene expression by *3.3. HPV mRNA Expression*

*3.3. HPV mRNA Expression* 

qPCR. Specifically, HPV16 E7 gene expression was analyzed in 16 HPV-positive HNSCC HPV-positive HNSCC samples were analyzed for HPV16 E7 gene expression by qPCR. Specifically, HPV16 E7 gene expression was analyzed in 16 HPV-positive HNSCC samples. mRNA E7 expression was detected in 15/16 (93.75%) (Figure 1C). Pearson correlation test showed no correlation between the expression levels (log10) of E7 and HPV DNA load (r = 0.42, *p* > 0.05). Furthermore, Spearman correlation analyses showed correlation between E7 expression and p16 up-regulation (r = 0.59; *p* < 0.05) (Figure 1D). But, HPV E7 mRNA expression did not correlate to p16 upregulation for all samples, since two samples (one OSCC and one OPSCC) presented E7 expression with p16 downregulation, and one

sample presented p16 upregulation but no E7 expression; therefore, p16 is not always a good marker of HPV infection. HPV E7 mRNA expression did not correlate to p16 upregulation for all samples, since two samples (one OSCC and one OPSCC) presented E7 expression with p16 downregulation, and one sample presented p16 upregulation but no E7 expression; therefore, p16 is not

samples. mRNA E7 expression was detected in 15/16 (93.75%) (Figure 1C). Pearson correlation test showed no correlation between the expression levels (log10) of E7 and HPV DNA load (r = 0.42, *p* > 0.05). Furthermore, Spearman correlation analyses showed correlation between E7 expression and p16 up-regulation (r = 0.59; *p* < 0.05) (Figure 1D). But,

#### *3.4. Serological Studies* always a good marker of HPV infection.

#### 3.4.1. HPV16 L1 Antibody Titer *3.4. Serological Studies*

*Cancers* **2021**, *13*, x 8 of 18

Serum from all HPV-positive (*n* = 20) HNSCC patients (Table 1) and from HPVnegative (*n* = 8) HNSCC patients, consisting of 4/8 (50%) OPSCCs and 4/8 (50%) OSCCs, were tested for HPV16 L1 IgG antibodies. HPV16 L1 antibodies were found with a similar proportion in 18/20 (90%) HPV–positive HNSCC and 7/8 (87.5%) HPV-negative HNSCC patients at T0 (*p* > 0.05). HPV DNA-positive HNSCC patients presented higher Optical Density (OD) readings for antibodies anti-HPV16 L1 compared to HPV-negative (Mean ± [SD], 4.001 ± 2.11 vs. 2.29 ± 0.32; *p* < 0.05) (Figure 2A). Antibody response was further compared during follow-up at 3, 6, 12 and 24 months. Results indicated that HPV16 L1 antibody titers did not vary significantly during follow-up (*p* > 0.05) (Figure 2B). 3.4.1. HPV16 L1 Antibody Titer Serum from all HPV-positive (*n* = 20) HNSCC patients (Table 1) and from HPV-negative (*n* = 8) HNSCC patients, consisting of 4/8 (50%) OPSCCs and 4/8 (50%) OSCCs, were tested for HPV16 L1 IgG antibodies. HPV16 L1 antibodies were found with a similar proportion in 18/20 (90%) HPV–positive HNSCC and 7/8 (87.5%) HPV-negative HNSCC patients at T0 (*p* > 0.05). HPV DNA-positive HNSCC patients presented higher Optical Density (OD) readings for antibodies anti-HPV16 L1 compared to HPV-negative (Mean ± [SD], 4.001 ± 2.11 vs. 2.29 ± 0.32; *p* < 0.05) (Figure 2A). Antibody response was further compared during follow-up at 3, 6, 12 and 24 months. Results indicated that HPV16 L1 antibody titers did not vary significantly during follow-up (*p* > 0.05) (Figure 2B).

**Figure 2.** ELISA tests on HNSCC serum samples. Statistical significance was indicated as \* for *p* < 0.05; (**A**) Serum antibody levels against HPV16 L1 in HNSCC patients. Differential OD between HPV-positive and HPV-negative patients (*p* < 0.05); (**B**) HPV16 L1 antibody variation during HPVpositive patient follow-up; (**C**) HPV16 E7 oncoprotein quantification in serum shows no difference between HPV-positive and HPV-negative patients (*p* > 0.05); (**D**) HPV16 E7 oncoprotein variation during HPV-positive patient follow-up.

#### 3.4.2. HPV16 E7 Oncoproteins in Sera

HPV16 E7 oncoprotein (ng/mL) amounts were evaluated at the time of diagnosis and during follow-up at 3, 6, 12 and 24 months. At T0, HPV16 E7 oncoprotein was detected in 6/20 (30%) HPV-positive patient serum and no HPV-negative cases (*p* > 0.05) (Figure 2C). Variation in the amount of E7 oncoprotein during follow-up was studied. Nine out of 20 (45%) patients showed an increment in the amount of oncoprotein during follow-up; 4/9 (44.44%) patients were positive at the time of diagnosis, while 5/9 (55.55%) became positive during follow-up. Two patients out of 20 (10%) positive at the time of diagnosis, presented HPV E7 decrement over-time, and one became negative. HPV16 E7 variation in samples during follow-up resulted statistically insignificant (*p* > 0.05) (Figure 2D). Nine out 20 (45%) patients were E7 negative at T0 and remained negative during follow-up.

Finally, HPV16 E7 oncoprotein levels in serum were studied in correlation to the viral mRNA expression in the tumor samples. Results showed correlation between the amount of HPV16 E7 mRNA expressed in the tumors and E7 oncoprotein in serum (r = 0.79, *p* < 0.01), suggesting that circulating E7 protein may be due to release from the tumor site.

#### *3.5. Survival Analysis*

3.5.1. RFS and OS in Correlation to HPV DNA, p16 Expression and HPV mRNA

The median follow-up time of this study was 24 months. Relapse free survival (RFS) and overall survival (OS) were assessed in HPV-positive HNSCC patients compared to HPV-negative cases. Different RFS rates were observed; 72.11% and 48.77% for HPVpositive and -negative, respectively (*p* > 0.05) (Figure 3A). Furthermore, OS was improved for HPV-positive patients; 88.89% compared to 52.08% in HPV-negative OPSCCs (*p* < 0.01) (Figure 3B).

To study the effect of p16 expression on survival rate, all HNSCC samples were subdivided into p16-over or –underexpression in the tumor sample. Log<sup>2</sup> fold change (FC) value (with fixed interval) was used as the cut-off criteria. High and low expression were considered when FC was greater than 1 (*n* = 13) or lower than −1 (*n* = 17), respectively [35]. RFS was 73.84%, in patients carrying p16 upregulation, compared to p16 downregulation, 48.12% (*p* > 0.05) (Figure 3C). OS was 100% in patients with higher p16 expression compared to 52.94% of patients with p16 downregulation (*p* < 0.01) (Figure 3D).

RFS and OS were also assessed for HPV E7 mRNA expression in HNSCCs samples. Samples were divided into expressing E7 oncogene (*n* = 15) and non (*n* = 26). Survival proportions indicated that RFS was 64.61% in patients positive for E7 mRNA, and 48.77% in patients HPV mRNA-negative (*p* > 0.05) (Figure 3E). OS was higher in patients carrying HPV E7 mRNA, 92.85%, compared to HPV mRNA-negative, 52.08% (*p* < 0.05) (Figure 3F).

**Figure 3.** Kaplan-Meier (KM) curves for RFS and OS in HNSCC; KM curves for (**A**) RFS and (**B**) OS for HPV DNA presence in HNSCC tumor samples; KM curves for (**C**) RFS and (**D**) OS for p16 overor under-expression in HNSCC samples; KM curves for (**E**) RFS and (**F**) OS for HPV E7 mRNA expression in HNSCC tumor samples. Statistical significance was indicated as *p* < 0.01 or *p* < 0.05. **Figure 3.** Kaplan-Meier (KM) curves for RFS and OS in HNSCC; KM curves for (**A**) RFS and (**B**) OS for HPV DNA presence in HNSCC tumor samples; KM curves for (**C**) RFS and (**D**) OS for p16 over- or under-expression in HNSCC samples; KM curves for (**E**) RFS and (**F**) OS for HPV E7 mRNA expression in HNSCC tumor samples. Statistical significance was indicated as *p* < 0.01 or *p* < 0.05.

#### 3.5.2. RFS and OS in Relation to Serum HPV16 L1 Antibodies

3.5.2. RFS and OS in Relation to Serum HPV16 L1 Antibodies The next step was to study the association between HPV infection serological markers, such as HPV16 L1 antibody, with patient's survival. No significant differences were observed for HPV16 L1 antibodies in RFS or OS for HPV-positive patients (*n* = 20) at the time of diagnosis and during follow up. RFS was 51.28% and 100% for HPV16 L1 antibody-positive (*n* = 18), for HPV16 L1 antibody-negative (*n* = 2) patients, respectively (*p* > 0.05) (Figure 4A). OS was also similar between HPV-positive patients and HPV16 L1 antibody positivity or negativity, at 63.64% and 100%, respectively (*p* > 0.05) (Figure 4B). The next step was to study the association between HPV infection serological markers, such as HPV16 L1 antibody, with patient's survival. No significant differences were observed for HPV16 L1 antibodies in RFS or OS for HPV-positive patients (*n* = 20) at the time of diagnosis and during follow up. RFS was 51.28% and 100% for HPV16 L1 antibody-positive (*n* = 18), for HPV16 L1 antibody-negative (*n* = 2) patients, respectively (*p* > 0.05) (Figure 4A). OS was also similar between HPV-positive patients and HPV16 L1 antibody positivity or negativity, at 63.64% and 100%, respectively (*p* > 0.05) (Figure 4B). Overall these results indicate that HPV16 L1 is a poor indicator of prognosis and since it is a cumulative marker of exposure, it may be used solely for epidemiological purposes.

Overall these results indicate that HPV16 L1 is a poor indicator of prognosis and since it is a cumulative marker of exposure, it may be used solely for epidemiological purposes.

**Figure 4.** Kaplan-Meier (KM) curves for serological tests representing RFS and OS in HNSCC patients for HPV16 L1 and OPSCC for E7 oncoprotein; KM of (**A**) RFS and (**B**) OS for HPV16 L1 in HNSCC patients; KM of (**C**) RFS and (**D**) OS for HPV E7 oncoprotein in serum from HPV-positive OPSCC patients; KM of (**E**) RFS and (**F**) OS for increment or decrement of E7 oncoprotein in serum from OPSCC patients during follow-up. Statistical significance was indicated as *p* < 0.05. **Figure 4.** Kaplan-Meier (KM) curves for serological tests representing RFS and OS in HNSCC patients for HPV16 L1 and OPSCC for E7 oncoprotein; KM of (**A**) RFS and (**B**) OS for HPV16 L1 in HNSCC patients; KM of (**C**) RFS and (**D**) OS for HPV E7 oncoprotein in serum from HPV-positive OPSCC patients; KM of (**E**) RFS and (**F**) OS for increment or decrement of E7 oncoprotein in serum from OPSCC patients during follow-up. Statistical significance was indicated as *p* < 0.05.

#### 3.5.3. RFS and OS in Relation to Serum HPV16 E7 Oncoprotein

3.5.3. RFS and OS in Relation to Serum HPV16 E7 Oncoprotein HPV16 E7 oncoprotein in serum was correlated to patients' clinicopathological features (Table 1). Interestingly, E7 oncoprotein in serum was strongly associated to recurrence in HNSCC patients (*p* < 0.0001) and in the OPSCC subgroup (*p* < 0.001). Statistical analyses on other HNSCC subtypes were not possible due to the small sample size (Table 1). RFS was 0% for HNSCC with E7 positivity compared to 90.9% for patients testing negative for E7 protein (*p* < 0.0001) (Figure 4C). OS was 100% and 50% in patients negative HPV16 E7 oncoprotein in serum was correlated to patients' clinicopathological features (Table 1). Interestingly, E7 oncoprotein in serum was strongly associated to recurrence in HNSCC patients (*p* < 0.0001) and in the OPSCC subgroup (*p* < 0.001). Statistical analyses on other HNSCC subtypes were not possible due to the small sample size (Table 1). RFS was 0% for HNSCC with E7 positivity compared to 90.9% for patients testing negative for E7 protein (*p* < 0.0001) (Figure 4C). OS was 100% and 50% in patients negative and positive for E7 oncoprotein, respectively (*p* < 0.01) (Figure 4D).

and positive for E7 oncoprotein, respectively (*p* < 0.01) (Figure 4D). The variation in serum E7 oncoprotein was also studied in correlation to patient survival. RFS was 42.85% in HNSCC patients who increased E7 oncoprotein amounts during follow-up, compared to 79.55% in those who experienced E7 decrease (*p* > 0.05) (Figure 4E). OS proportion was 85.71% for patients showing increased E7 oncoprotein, and

90.9% for those showing a decreased E7 oncoprotein (*p* > 0.05) (Figure 4F). These results highlight the importance of patient monitoring for recurrence after circulating HPV E7 oncoprotein being found at the time of diagnosis or increasing levels during follow-up. for those showing a decreased E7 oncoprotein (*p* > 0.05) (Figure 4F). These results highlight the importance of patient monitoring for recurrence after circulating HPV E7 oncoprotein being found at the time of diagnosis or increasing levels during follow-up.

The variation in serum E7 oncoprotein was also studied in correlation to patient survival. RFS was 42.85% in HNSCC patients who increased E7 oncoprotein amounts during follow-up, compared to 79.55% in those who experienced E7 decrease (*p* > 0.05) (Figure 4E). OS proportion was 85.71% for patients showing increased E7 oncoprotein, and 90.9%

#### *3.6. TNM Stage in Correlation to OPSCC Patient Prognosis and E7 Oncoprotein in Serum 3.6. TNM Stage in Correlation to OPSCC Patient Prognosis and E7 Oncoprotein in Serum*

*Cancers* **2021**, *13*, x 12 of 18

RFS was 72.9% and 57.14% for patients with T (1–2) and T (3–4), respectively (*p* > 0.05) (Figure 5A), while OS was 87.5% and 90%, respectively (*p* > 0.05) (Figure 5B). Similarly, no statistically significant differences were observed for RFS or OS survival rates when studied in correlation to lymph node involvement; RFS was 100% vs. 58.18% for patients without and with lymph node involvement (*p* > 0.05) (Figure 5C), while OS was similar; 100% and 86.67%, respectively (*p* > 0.05) (Figure 5D). RFS was 72.9% and 57.14% for patients with T (1–2) and T (3–4), respectively (*p* > 0.05) (Figure 5A), while OS was 87.5% and 90%, respectively (*p* > 0.05) (Figure 5B). Similarly, no statistically significant differences were observed for RFS or OS survival rates when studied in correlation to lymph node involvement; RFS was 100% vs. 58.18% for patients without and with lymph node involvement (*p* > 0.05) (Figure 5C), while OS was similar; 100% and 86.67%, respectively (*p* > 0.05) (Figure 5D).

**Figure 5.** Kaplan-Meier (KM) curves for tumor size (T), node status (N) and stage in HPV-positive HNSCC patients representing RFS and OS; KM representing (**A**) RFS and (**B**) OS for patients divided into tumor size: T (1–2) and T (3–4); KM representing (**C**) RFS and (**D**) OS for patients divided into node status: N0 and N+; KM representing (**E**) RFS and (**F**) OS for patients divided into stages I/II and III/IV; OS for patients divided into stages I/II and III/IV. Statistical significance was indicated as *p* < 0.05.

Patients in stages III/IV are more likely to recur. Indeed, RFS for patients in stage III/IV was 42.86% compared to 100% for patients in stage I/II (*p* < 0.05) (Figure 5E), while OS was similar for both groups 83.33% and 100%, respectively (*p* > 0.05) (Figure 5F).

Lastly, we studied the correlation between serum E7 presence and tumor size, lymph node involvement and disease stage. E7 in serum correlated to tumor size (*p* < 0.05), but not to lymph node involvement in OPSCC (*p* > 0.05) (Table 1). Out of 6 HPV-positive patients with stage I/II, none presented E7 oncoprotein in serum at T0, while 5/12 (41.66%) of the patients in stage III/IV presented E7 oncoprotein in serum (*p* > 0.05).

#### **4. Discussion**

The current study aims to find markers for recurrence in HPV-positive patients. For patient stratification, we studied classical HPV markers, such as HPV DNA, p16 mRNA and viral mRNA expression. Once stratified, we studied the presence of potential serological markers, i.e., HPV16 L1 antibodies and, for the first time, the HPV E7 oncoprotein. Serological markers were then correlated to patient prognosis.

In an initial screening, we found that 29.85% of HNSCC tumor samples, including 75% OPSCC, harbored HPV-DNA, and 95% tested HPV16-positive. These findings are in accordance to other studies indicating that HPV is found in 25% of all HNSCCs, and in up to 70% of OPSCC tumors [36–39], whereas 90% of all HPV-positive tumors carried HPV type 16 [40]. P16 mRNA expression was found to be overexpressed in 75% of HPV-positive HNSCC samples. P16 is an established surrogate marker for tumors with transcriptionally active HPV, which is known to be associated with tumors that respond better to therapy and have improved outcomes [41,42]. Yet, not all HPV-positive tumors show p16 gene upregulation, as shown in herein and in previous studies [15]. Transcription of E7 viral oncogene was assessed in HPV-DNA positive patients only, showing 93.75% of HPV16 DNA positive samples expressing the HPV16 E7 oncogene, in agreement with previous studies [30].

Since HPV status has a great impact on patient prognosis for different HNSCCs, such as OPSCC and OSCC [43,44], it is important to stratify patients correctly. In our study, we found that 30% of HPV-positive patients presented recurrence within the first two years of diagnosis, similarly to other studies [9,11–13]. Classical HPV infection markers, i.e., HPV DNA, p16 and HPV mRNA showed improved patient OS in our cohort of study, but none of them correlated with recurrence.

Antibody response against L1 was studied for the prevalence of viral infection in HNSCC patients. HPV L1 antibodies are cumulative markers of past and present infection, although their presence does not imply HPV-driven tumorigenesis [45]. Indeed, in our study both HPV-positive and -negative HNSCC patients had antibodies against HPV16 L1 with a similar level of prevalence at 90% and 87.5%, respectively. Interestingly, the antibody titer in HPV-positive patients was higher compared to HPV-negative cases, which could be indicative of active infection.

We also studied the antibody response during the follow-up to monitor disease status, as proposed by Routman et al. [46], but no significant antibody titers change was observed during follow up, indicating that antibody levels against HPV L1 may not be useful in diagnosing or monitoring the disease.

Recently, the study of antibody response against HPV E6/E7 oncoproteins in OPSCC, the major subtype of HNSCC, has been proposed as a marker for disease progression. In spite of good perspectives for both diagnosis and prognosis [19,20,22,23,47], results are still under debate due to the lack of seroconversion in many patients, as was also underlined before for other diseases which are related to viral infections [48]. In a study conducted by Kreimer et al., 57.6% of OPSCC patients remained HPV E6 seronegative during follow-up [47].

To our knowledge, no previous research has been conducted to detect HPV E7 oncoprotein in HNSCC patient serum, while the availability of ELISA kits for oncoprotein detection could rapidly facilitate such study outcomes into clinical use. E7 oncoprotein in

serum was specifically found with a frequency of 30% in HPV-positive samples, while the detection of circulating protein at the time of diagnosis strongly correlated to recurrence. Our data are in accordance to other studies showing higher levels of antibodies against HPV oncoproteins at the time of diagnosis in association to a significantly increased risk of recurrence [22,49]. Similar to previous studies on antibody titer variation during follow-up, our investigation found no variation, increment or decrement of E7 oncoprotein in OPSCC serum could be association with patient outcome during the two-year follow-up [20,21,50]. Nevertheless, an increase or decrease in serum E7 during follow-up was observed in patients whether experiencing recurrence or not, respectively, thus, lack of significant correlation between serum E7 level and relapse may be due to limited sample sizes.

It is to be noted that, circulating E7 protein showed correlation with high E7 mRNA expression in the tumor, suggesting that tumor sites may provide the circulating oncoprotein. Circulating E7 protein may be considered a tumorigenesis marker, representing at serological level what occurs at the tumor site. Sources of viral oncoproteins in serum have currently not been established, but some hypotheses could be proposed. Firstly, the transcriptionally active circulating tumor cells may account for the presence of serum viral oncoproteins [13]. Indeed, HPV spreading through blood cells has been previously reported [51], while HPV E6/E7 transcription in circulating tumor cells (CTCs) has been correlated to patient prognosis [13]. Secondly, invasion and the associated development of a tumor vascular bed may result in the release of E6/E7 proteins from the tumor mass, probably as a consequence of necrosis [52]. Thirdly, HPV-positive tumor cells may secrete exosomes containing viral oncoproteins, as has been reported for other DNA viruses [53]. Whatever the mechanism, HPV16 E7 oncoproteins were successfully found in HNSCC patient serum and correlated to patient prognosis. E7 oncoprotein detection in serum at the time of diagnosis displayed strong diagnostic and prognostic reliability in predicting relapses and overall survival in HPV-positive HNSCC patients, especially HPV-positive OPSCC patients. Moreover, since HPV mRNA may be present in HPV-DNA negative samples [54], the analysis of HPV mRNA should be taken into consideration in all HNSCCs to avoid HPV-driven tumors misclassification.

Moreover, E7 oncoprotein also correlated to tumor size, but not lymph node involvement or disease stage. Overall, 41.66% HNSCC patients with high disease stage III/IV presented E7 oncoprotein in serum, while none of those with low stage I/II did so, in agreement with previous serologic studies [20,55], making the detection of E7 oncoprotein in serum an excellent discriminator for HNSCC patients that may relapse, especially for OPSCC patients.

This study demonstrates for the first time, the presence of circulating E7 oncoproteins in serum from HNSCC patients using a direct ELISA assay. Our results indicate that the presence of E7 oncoprotein in OPSCC patient serum at the time of diagnosis is indicative of a higher risk of recurrence. Liquid biopsy for the detection of prognostic markers in HPV-positive OPSCC patients provides valuable information on disease progression and may help stratify and monitor patients over-time; this, can result extremely useful for patients presenting persistent or occult tumors. For future studies, in order to increase the statistical power of the study, a larger sample size for all HNSCC subtypes will be considered. This study takes medicine one step closer to correct patient stratification for therapy de-intensification. The combination of classical markers with serological markers, may be used to plan personalized treatment strategies for HPV-positive patients.

#### **5. Conclusions**

Detection of circulating HPV E7 oncoprotein at the time of diagnosis, may be used as non-invasive procedure for patient stratification and follow-up, ultimately providing a tool for clinicians to determine which patients would be good candidates for treatment de-escalation or should be kept under close surveillance.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/cancers13133370/s1, Table S1 Validated primer sets used in qPCR to detect and quantify HPV DNA and both, viral and cellular genes.

**Author Contributions:** Conceptualization: F.M. and S.P.; methodology, L.O.-G. and J.C.R.; software, E.M.; validation, M.T., S.P., F.M.; formal analysis, L.O.-G., J.C.R.; investigation, L.O.-G., J.C.R., I.B., M.R.I.; resources, L.C., N.M., A.C., C.B.; supervision S.P.; data curation L.O.-G., J.C.R., C.L., I.B.; writing—original draft preparation, L.O.-G.; writing—review and editing, J.C.R., C.L., M.T., S.P., F.M.; visualization, L.O.-G.; supervision, M.T., S.P., F.M.; project administration, M.T., S.P., F.M.; funding acquisition, J.C.R., M.T., F.M. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was partially supported by grant IG 21956 (awarded to John Charles Rotondo) and by grant IG 21617 (awarded to Mauro Tognon) from the Associazione Italiana per la Ricerca sul Cancro (AIRC), Milan, Italy.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of Ferrara. (Authorization *n*. 160986, 12 December 2016).

**Informed Consent Statement:** Informed written consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Acknowledgments:** The pGEM1\_HPV45/pUC19\_HPV52 plasmids, which are commercially available, were a kind gift from Massimo Tommasino, Infectious and Cancer Biology group, International Agency for Research on Cancer (IARC), Lyon, France.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **References**

