Identification of transcriptionally active HPV infection in formalin-fixed, paraffin-embedded biopsies of oropharyngeal carcinoma

Human Pathology (2015) xx, xxx–xxx

www.elsevier.com/locate/humpath

Original contribution

Identification of transcriptionally active HPV infection in formalin-fixed, paraffin-embedded biopsies of oropharyngeal carcinoma☆ Patrizia Morbini MD a,b,⁎, Paola Alberizzi BS b , Carmine Tinelli MD c , Chiara Paglino MD d , Giulia Bertino MD e , Patrizia Comoli MD f , Paolo Pedrazzoli MD d , Marco Benazzo MD e a

Department of Molecular Medicine, Unit of Pathology, University of Pavia, via Forlanini 16, 27100 Pavia, Italy Department of Pathology IRCCS Policlinico S. Matteo Foundation, via Forlanini 16, 27100 Pavia, Italy c Clinical Epidemiology and Biometric Unit, IRCCS Policlinico S. Matteo Foundation, P.le Golgi 5, 27100 Pavia, Italy d Department of Oncology, IRCCS Policlinico S. Matteo Foundation, P.le Golgi 5, 27100 Pavia, Italy e Department of Otolaryngology, University of Pavia and IRCCS Policlinico S. Matteo Foundation, P.le Golgi 5, 27100 Pavia, Italy f Pediatric Hematology/Oncology, IRCCS Policlinico S. Matteo Foundation, P.le Golgi 5, 27100 Pavia, Italy b

Received 14 October 2014; revised 4 December 2014; accepted 25 December 2014

Keywords: Human papillomavirus; Oropharyngeal cancer; In situ hybridization; p16; mRNA in situ hybridization

Summary Human papillomavirus (HPV) oncogenic activity is the result of viral oncogene E6 and E7 expression in infected cells. Oncogene expression analysis is, however, not part of the routine diagnostic evaluation of HPV-associated oropharyngeal squamous cell carcinoma (OPSCC) since it requires fresh tumor tissue. We compared the diagnostic accuracy of several methods commonly employed for HPV characterization in OPSCC with the results of the newly available HPV E6/E7 mRNA in situ hybridization (ISH) on formalin-fixed, paraffin-embedded biopsy samples, in order to establish if the latter should be introduced in the diagnostic routine to increase accuracy when fresh tissue is not available. p16 immunostain, DNA ISH for high-risk HPV genotypes, SPF LiPA amplification and genotyping, and HPV16 E6 amplification were performed on 41 consecutive OPSCC samples. Twenty (48.7%) cases were positive by mRNA ISH; sensitivity and specificity were 100% and 90% for p16, 90% and 100% for DNA ISH, 70% and 76% for SPF10 LiPA, 90% and 76% for E6 amplification. A diagnostic algorithm considering p16 immunostain as first step followed by either high-risk HPV DNA ISH or HPV16 E6 amplification in p16-positive cases correctly characterized 90% of mRNA-positive and all mRNA-negative cases; combining the 3 tests correctly identified all cases. While no stand-alone test was sufficiently accurate for classifying HPV-associated OPSCC, the high sensitivity and specificity of the established combination of p16 immunostain, DNA ISH, and HPV16 DNA amplification suggests that the introduction of labour- and cost-intensive mRNA ISH, is not necessary in the diagnostic routine of oropharyngeal tumors. © 2015 Elsevier Inc. All rights reserved.

☆

Dedicated to the memory of Dr Barbara Dal Bello. ⁎ Corresponding author. Department of Molecular Medicine, Unit of Pathology, Via Forlanini 16, 27100 Pavia, Italy. E-mail addresses: [email protected] (P. Morbini), [email protected] (P. Alberizzi), [email protected] (C. Tinelli), [email protected] (C. Paglino), [email protected] (G. Bertino), [email protected] (P. Comoli), [email protected] (P. Pedrazzoli), [email protected] (M. Benazzo). http://dx.doi.org/10.1016/j.humpath.2014.12.014 0046-8177/© 2015 Elsevier Inc. All rights reserved.

2

1. Introduction Human papillomavirus (HPV)–associated carcinomas are a distinct subset of squamous cell carcinomas (SCC) arising from the oropharynx, whose number has significantly increased worldwide over the last two decades [1–3]. HPV-related oropharyngeal SCC (OPSCC) have clinical, biological, and histopathological characters different from conventional tobacco-related, HPV-negative SCC, and usually show better survival outcome [1,4,5] and response to chemoradiation [6–8]. OPSCC's less aggressive behaviour justifies the ongoing trials assessing the effect of therapy de-intensification in patients with HPV-associated oropharyngeal cancers [9]. Defining the criteria that should be used to select patients with HPV-induced tumors that are eligible for therapy de-escalation has become a critical issue for head and neck pathologists and oncologists. Several methods are available to assess the presence of HPV infection in tumor tissues, the most widely available being the identification of HPV DNA either by polymerase chain reaction (PCR) or in situ hybridization (ISH) [10]. However, the presence of viral DNA in a tumor sample can not be considered as definite evidence of virus-promoted neoplastic transformation [11], as the last requires active transcription of E6 and E7 viral oncogenes [12]; indirect evidence of viral oncogene transcription can be obtained with immunohistochemical stain for p16INK4a (p16), a protein which is normally upregulated as a consequence of HPV E7 protein expression and interaction with the cellular oncosuppressor pRB [12]. Several diagnostic algorithms have been proposed for the identification of HPV-associated OPSCC, most of which pair DNA analysis with immunohistochemistry [13–15]. Both immunohistochemistry and DNA-based tests have repeatedly been validated against the gold standard of E6 and E7 mRNA expression assessed with reverse-transcriptase (RT) PCR [14,16]. Although most validation studies reported good sensitivity and specificity [14,16–19], in all series HPV oncogene mRNA was amplified in few cases negative with DNA-based studies and/or p16 immunostain. Gold-standard mRNA amplification can hardly become the routine diagnostic test for HPV-associated OPSCC, since it requires fresh-frozen tissue and dedicated facilities for mRNA extraction and reverse transcription. A new ISH technique using mRNA probes for HPV E6/E7 gene transcripts has recently been introduced for formalin-fixed and paraffin-embedded (FFPE) tissues [20], thus overcoming the limitations of mRNA extraction from tumor samples. This test showed a strong correlation with mRNA RT-PCR with very few discordant results [21–23]. The present study compared the diagnostic accuracy of several widely employed analysis methods for p16 expression and HPV DNA identification with the in situ determination of HPV oncogene mRNA in oropharyngeal tumor samples, in order to establish if mRNA ISH can

P. Morbini et al. identify cases that would remain unrecognized with current methods and thus should be introduced in the diagnostic routine to increase accuracy when only FFPE samples are available.

2. Materials and methods 2.1. Patient selection Forty-one consecutive patients referring to the Otolaryngology Clinic of our Institution from January 2010 to June 2013 with SCC originating from the oropharyngeal area (palatine tonsil, base of tongue, soft palate) were enrolled in the study. Biopsy or surgical samples obtained from each patient were fixed in formalin and processed routinely for histopathological study. The protocol was reviewed and approved by the Institutional Ethical Review Board and is in compliance with the Helsinki Declaration. Each subject enrolled in the project signed a detailed informed consent form. Hematoxylin and eosin–stained sections were reviewed to assess the presence of keratinization in tumor cells.

2.2. p16 Immunostain For the immunohistochemical detection of p16, the CINtec Histology Kit (MTM Laboratories AG, Heidelberg, Germany) was used according to the manufacturer’s instructions on a Ventana BenchmarkTX automated stainer (Ventana Medical Systems Inc, Tucson, AZ). Samples of high-grade cervical intraepithelial lesions were used as positive controls. Cases were read by the study head and neck pathologist (P.M.). Strong and diffuse nuclear and cytoplasmic staining in N70% of the tumor cells was considered as positive according to previous studies [15,21].

2.3. HPV DNA in situ hybridization DNA ISH was performed on paraffin tissue sections with the INFORM-HPV III family 16 probe using the ISH I View Blue Plus Detection Kit according to the manufacturer's instructions, on a Ventana BenchmarkTX automated stainer. The probe hybridizes with high-risk (HR) HPV genotypes 16, 18, 33, 35, 45, 51, 52 56, and 66. Cases were read by the study head and neck pathologist (P.M.). Any definitive (diffuse or dot-like) nuclear staining in the tumor cells was considered positive. Cases were classified in a binary manner as either positive or negative.

2.4. HR HPV mRNA ISH RNA ISH for E6 and E7 mRNA was performed using the RNAscope assay according to manufacturer's instructions (Advanced Cell Diagnostics, Hayward, CA). Briefly, 4-mm-thick unstained tissue sections were pretreated with

HPV in FFPE oropharyngeal carcinoma biopsies heat and protease. Target probes to DapB (negative control), PoliA (positive control), HPV16, and HR HPV cocktail (including types 18, 31, 33, 35, 52, and 58) were separately hybridized to adjacent tissue sections. Tumors that tested positive for HR HPV genotypes other than 16 (HPV33, HPV3, HPV52) with DNA analysis were also hybridized with the type-specific target probes. Hybridization signals were amplified by sequential hybridization with preamplifier, amplifier, and label probes and then visualized by chromogenic staining with either fast red or diaminobenzidine. Stains were reviewed by the study pathologist at 20× magnification. Staining above any staining that was present on the DapB negative control was considered positive [23]. Cases with high background staining on the DapB negative control or with lack of staining on the PoliA positive control were deemed technical failures and the tests were repeated.

3 generated for each test. The univariate analysis for diseasefree (DFS) and disease-specific survival (DSS) was performed by computing survival curves according to the Kaplan-Meier method and comparisons for prognostic factors were evaluated by the log-rank test. Factors tested for prognostic value included sex, age, tumor size (T, according to the TNM Classification of the UICC 2010), incidence and extent of lymph node metastasis (N, according to the TNM Classification of the UICC 2012), tobacco and alcohol consumption, primary tumor site, therapy (isolated tumor or nodal surgery, chemotherapy or radiotherapy, combined chemo and radiotherapy, trimodal treatment), keratinizing histology, and HPV status in the tissue samples. P b .05 were considered to be statistically significant and all tests were two-side. Data analysis was performed with STATA statistical package (release 11.1, 2010, Stata Corporation, College Station, TX).

2.5. HPV DNA amplification and typing DNA extraction and HPV typing of biopsy samples were performed according to previously reported protocols, using the line probe INNO-LiPA HPV genotyping assay version Extra (Fujirebio Europe, Ghent, Belgium). The Extra version of the assay allows the simultaneous and separate detection of 18 HR (16, 18, 26, 31, 33, 35, 39, 45, 51, 52, 53, 56, 58, 59, 66, 68, 73, and 82), 7 LR (6, 11, 40, 43, 44, 54, and 70), and 2 unclassified HPV types (69/71 and 74). Hybridization patterns were analysed with the specific scanner Line reader and Analysis Software for LiPA HPV, and the results were confirmed by two independent readers.

2.6. HPV16 E6 amplification The DNA extracted from FFPE samples was amplified using a primer pair (3′: CCATGCATGATTACAGCTGG; 5′: GAACAGCAATACAACAAACCG) that amplifies a 201-bp fragment within the E6 gene of HPV 16 [24]. Amplification parameters were 94°C for 10 minutes, followed by 95°C for 30 seconds, 60°C for 15 seconds, 72°C for 30 seconds for 35 cycles, and 72°C for 5 minutes for a final extension. Amplified products were analyzed by electrophoresis on 2% agarose gel. SiHA cells containing a known number of HPV16 DNA copies and blank reagents were used respectively as positive and negative controls in each amplification run.

2.7. Statistical analysis The Fisher exact test was used for comparison of demographic and tumor-specific parameters between HPV mRNA ISH-positive and -negative groups and for the correlation between other HPV test methods and the reference test. Analytical sensitivity and specificity against the reference test were calculated for all single diagnostic tests. Positive and negative predictive values were also

3. Results 3.1. Cohort data In 20 cases (48.8%), ISH showed the presence of mRNA of HR HPV genotypes in tumor cells. The characteristics of the entire cohort and of the HPV-positive and negative groups as classified with mRNA ISH are detailed in Table 1. HPV-positive patients had significantly less exposure to tobacco smoke (P = .01). HPV-associated tumors originated more frequently from the tonsil and the base of tongue, while only 1 HPV-positive case originated from the soft palate (P = .019). There were no statistical differences between the 2 groups for age, sex, alcohol consumption, and tumor and nodal stage. The median patient follow-up was 15 months (IQR, 8-22).

3.2. p16 and DNA-based test results Positive p16 stain was observed in 22 cases (53.7%). HR HPV DNA was found by SPF10 LiPA in 19 cases (46.3%). In 9 further cases, the test could not identify a specific viral type. HPV16 was identified by SP10 LiPA in 17 of the 19 HR-positive tumors (89.5%). The other cases hosted HPV33 and 52. Coinfection by HR HPV16 and 35 was observed in one case. HPV16 E6 DNA was amplified in 23 cases (56.1%). DNA ISH was positive in 18 cases (43.9%); 15 cases showed only integrated hybridization signals; in 3 cases a limited number of differentiated squamous cells displayed an episomal pattern of staining (Fig. 1D).

3.3. Comparison of test results 3.3.1. mRNA ISH-positive cases All 20 mRNA ISH-positive cases strongly expressed p16. Eleven (60%) were positive with all DNA-based tests and

4

P. Morbini et al. Table 1

Patient and tumor cohort data Total (%)

No. of patients Mean age at diagnosis (SD) Sex Female Male Tumor site Tonsil Soft palate Base of tongue Nodal stage N0 N1 N2/3 TNM stage 1 2 3 4 Smoking Non-smoker Former smoker (N5 y) Current smoker Alcohol consumption Non-drinker b28 U/wk N28 U/wk Keratinizing histology Therapy Surgery (T) Surgery (N) Chemotherapy Radiotherapy

HPV status by HR HPV mRNA ISH

Statistical significance P

Positive (%)

Negative(%)

41 63.68 (10.14)

20 62.5

21 64.7

7 (17) 34 (83)

4 (20) 16 (80)

3 (14.2) 18 (85.7)

21 (51.2) 10 (24.3) 10 (24.3)

13 (65) 1 (5) 6 (30)

8 (38) 9 (42.8) 4 (19)

8 (19.5) 11 (26.8) 22 (53.6)

2 (10) 6 (30) 12 (60)

6 (28.5) 5 (23.8) 10 (47.6)

2 2 25 12

(4.8) (4.8) (60.9) (29.2)

1 (0.5) 0 13 (65) 6 (30)

1 (4.7) 2 (9.4) 12 (57.1) 6 (28.5)

11 (26.8) 7 (17) 22 (53.6)

9 (45) 4 (20) 6 (30)

2 (9.4) 3 (14.2) 16 (76.1)

NS NS

0.019

NS

NS

0.01

NS 22 10 9 16

(53.6) (24.4) (21.9) (39)

14 4 2 3

(70) (20) (10) (15)

8 (38) 6 (28.6) 7 (33.3) 13 (61.9)

15 18 35 36

(36.5) (43.9) (85.3) (87.8)

9 11 18 19

(45) (55) (90) (95)

6 (28.5) 7 (33.3) 17 (80.9) 17 (80.9)

0.003 NS

Abbreviations: HR, high risk; HPV, human papillomavirus; ISH, in situ hybridization; NS, nonsignificant.

HPV16 was exclusively documented; 5 (25%) were negative for HR HPV DNA with SPF10 LiPA, although in 2 of them uncharacterized viral DNA was documented; 2 (10%) were negative for DNA ISH and hosted, respectively, HPV 16 and 35 and an uncharacterized genotype with SPF10 LiPA. HPV16 E6 amplification was positive in all the above cases but one (SPF10 LiPA-negative, DNA ISH-positive). HR HPV genotypes other than 16 were identified by SPF10 LiPA in 3 (7.3%) cases (HPV52, 33, and 35). In two, a coinfection with HPV16 was documented (one by LiPA, one by HPV16 E6 amplification); the last case had an isolated HPV33 infection. Type-specific mRNA ISH documented exclusive HPV16 transcription in the HPV16-52 coinfection (Fig. 1A-F), transcription of both HPV16 and 35 in the HPV16-35 coinfection (Fig. 1G-M), and transcriptional activity of HPV33 but not of HPV16 in the case where only HPV33 DNA was identified (Fig. 2A-F). The mRNA and DNA ISH-positive case with an uncharacterized infection by SPF10 LiPA that was negative for HPV16 E6 amplification was also negative for mRNA ISH

with HPV16-specific probes, suggesting a non-HPV16 HR infection (Fig. 2G-N). mRNA ISH-positive cases with discordant test results are summarized in Table 2. 3.3.2. mRNA ISH-negative cases Among the 21 mRNA ISH-negative cases, 10 were negative with all other tests. SPF10 LiPA identified HPV16 DNA in 5 (23.8%) and uncharacterized viral DNA in an other 5 (23.8%). HPV16 E6 amplification confirmed the presence of HPV16 DNA in all SPF10 LiPA HPV16-positive cases, while the uncharacterized infections were negative. Two cases (9.5%), one hosting an uncharacterized infection by SPF10 LiPA, the other negative with all DNA-based tests, had positive p16 immunostain. DNA ISH was negative in all 21 cases. mRNA ISH-negative cases with discordant test results are summarized in Table 2. 3.3.3. Statistical analysis Fisher exact test showed highly significant correlation with HR HPV mRNA ISH results for all tests. A strong

HPV in FFPE oropharyngeal carcinoma biopsies

5

Fig. 1 Transcriptional activity and marker profile in OPSC coinfected by HPV16 and other HR genotypes. Non-keratinizing SCC of the tongue base, with maturation (A) showing diffuse p16 immunostain (B) and positive HR HPV DNA with both integrated (C) and episomal pattern (D). Although HPV16 and 52 DNA were amplified, only HPV16 mRNA was positive with ISH (E), while HPV52 was negative (F). Non-keratinizing tonsillar SCC (G), showing diffuse p16 immunostain (H) but negative HR HPV DNA ISH (I). The tumor was positive for HR HPV (J), HPV16 (K), and HPV35 (L) mRNA ISH. Both HPV16 and 35 DNA were amplified from the tumor; original magnification: A, B, and H, 20×; C, D, I, and J, 40×; E, F, K, and L, 100×; G, 10×.

correlation was also observed between HR HPV and HPV16 mRNA ISH (Table 3). Sensitivity and specificity of each test with respect to mRNA ISH are detailed in Table 2. p16 immunostain showed 100% sensitivity for transcriptionally active HPV infection, while HR HPV DNA ISH provided 100% specificity.

3.4. Diagnostic strategy analysis The 100% negative predictive value of p16 immunostain confirmed it as the test of choice for initial HPV screening of OPSCC. Performing either HR HPV DNA ISH or HPV16 E6

amplification in p16-positive cases allowed the correct definition of 90% mRNA-positive cases and of both p16-positive, mRNA-negative cases. By pairing p16 with HR HPV DNA detection by SPF10 LiPA, 70% of mRNA-positive cases and both mRNA-negative cases were recognized. HPV16 E6 DNA amplification as a third step in p16-positive, DNA ISH-negative cases, allowed the correct classification of all 4 cases with a p16-positive/DNA ISH-negative profile (2 positive for HR HPV mRNA and 2 negative). The reverse (DNA ISH as a third step in p16-positive/HPV16 E6 DNA-negative cases) provided the same result.

6

P. Morbini et al.

Fig. 2 Transcriptional activity and marker profile in OPSCC infected by non-16 HR HPV genotypes. Non-keratinizing tonsillar SCCs (A and G), showing diffuse p16 immunostain (B and H) and positive HR HPV DNA ISH (C and I). The first case was positive for HPV33 mRNA ISH (D and E) and negative for HPV16 mRNA ISH (F). Only HPV33 DNA was amplified from the tumor. The second case was positive for HR HPV mRNA ISH (J and K) and negative for HPV16 mRNA ISH (L). No specific HPV genotype was identified by amplification studies; original magnification: A, 20×; B, D, G, and J, 40×; H, 10×; C, E, F, I, K, and L, 100×.

3.5. Survival analysis The patients were followed for a median of 15 months (IQR, 8-22; range, 4-41 months). There was no difference in the number of patients that experienced tumor persistence, recurrence, and tumor-related death in mRNA-positive and negative groups (Table 4). In univariate analysis, DFS was significantly associated with p16 expression (P = .022) and positive DNA ISH (p: 0.038), while positive mRNA ISH was just above the significance threshold (P = .063). Age, nodal neck dissection, and the administration of combined radio-chemotherapy, independently of surgical treatment, were also significantly associated with DFS (P = .005, .027, and .009 respectively). DSS showed a significant association

with age and a therapeutic approach combining surgery, chemotherapy, and radiotherapy (P = .011 and .044, respectively), while a moderate, non-significant association was observed with p16 expression (P = .076) and positive DNA ISH (P = .069).

4. Discussion Clinical trials aimed at assessing the outcomes and benefits of de-escalated therapies in patients with HPV-driven OPSCC require the definition of unequivocal criteria to distinguish these patients from those with tumors induced by

HPV in FFPE oropharyngeal carcinoma biopsies

7 thus be introduced in the diagnostic routine. In our series of 41 consecutive patients, 20 expressed HR HPV mRNA, and all of them were classified as strongly positive for p16. This result further confirms that p16 expression has the highest specificity for HPV transcriptional activity [19,23,25,26], while its sensitivity is limited by recurrent p16-positive cases that are negative with all DNA and mRNA-based tests. With methods that identify a large spectrum of HR HPV genotypes, such as INFORM ISH or SPF10 LiPA, we confirmed that HR genotypes other than 16 can be involved in oropharyngeal oncogenesis in a proportion (10%) similar to that reported by other mRNA-based studies [22] and could account for at least some discrepant p16-positive results observed when HPV16 was the only investigated genotype [14]. Interestingly, we also documented the concomitant transcription of HPV16 and 35 E7 mRNA, while in a sample coinfected by HPV16 and 52 only the first one was active. Only 2 (9%) p16-positive cases in our series were negative for HR HPV by all tested methods. In these cases other mechanisms of p16 activation, such as inactivating mutations in the retinoblastoma protein pathway [27] should be investigated to explain the protein overexpression. As previously reported [28], inactive HPV16 infections were frequent in OPSCC, all of which were negative for p16. More surprisingly, more than one third (35%) of cases with a transcriptionally active HPV16 infection were not recognized by SPF10 amplification and reverse linear array, which failed to identify HPV DNA (3 cases), to characterize the genotype (3 cases) or, in one case, to identify HPV16 coinfection with another characterized HR genotype. Other studies reported similar results, albeit in a lower proportion of cases [19]. Most of these failures can be explained when considering SPF10 target region, which encompasses 65 bp within the L1 HPV gene. Since this region is lost upon viral integration in the host cell genome, when HPV is largely or entirely integrated the amount of target DNA will possibly fall below the detection threshold of SPF10 or other consensus primers targeting the L1 region [29,30]. Confirming this observation, a recent study documented by quantitative PCR that the ratio of episomal versus integrated viral copies was lower in cervical samples showing an exclusive punctuate DNA ISH signal [31], which was the predominant pattern observed in our series. Primers that

Table 2 Detail of immunohistochemical and molecular findings in discordant cases mRNA ISH

p16

DNA ISH

SPF10LiPA

HPV16 E6

Type-specific mRNA ISH

Positive 1 2 3 4 5 6 7 8 9

+ + + + + + + + +

+ + + + + − + − +

− − − NT NT NT 52 16,35 33

+ − + + + + + + −

16 − 16 16 16 16 16 16,35 33

Negative 1 2 3 4 5 6 7 8 9 10 11

− + − − − − − − − + −

− − − − − − − − − − −

NT NT NT 16 16 16 16 NT NT − 16

− − − + + + + − − − +

− − − − − − − − − − −

Abbreviations: NT, non-specific genotype.

tobacco smoke and alcohol. The presence of viral oncogene mRNA in tumor cells has been proposed as the gold standard for the identification of HPV-driven neoplastic transformation, but none of the trials that have been initiated so far are relying on this criterion for patient selection [9]. Indeed, mRNA analysis would be hardly available in the diagnostic routine of most pathology laboratories, which are more familiar with in situ determination of biological markers on tissue sections. We determined the transcriptional activity of HR HPV genotypes in a series of OPSCC, using a new in situ method that has demonstrated high sensitivity and specificity for the recognition of specific messengers as compared with RT-PCR [22], and compared its results with those obtained with established laboratory methods, in order to assess if mRNA ISH helps to identify misclassified cases and should

Table 3

Diagnostic performance of HPV testing methods as compared with HR HPV mRNA ISH

Test

Number positive (%)

Sensitivity

Specificity

PPV

NPV

P

Non-keratinizing histology p16 immunostain HR HPV DNA by SPF10 LiPA HPV16 by SPF10 LiPA HPV16 E6 DNA HR HPV DNA ISH HPV16 mRNA ISH

20 (48.7) 22 (53.6) 19 (46.3) 17 (41.4) 24 (58.3) 18 (43.9) 18 (43.9)

0.85 1 0.70 0.76 0.90 0.90 0,90

0.62 0.9 0.76 0.60 0.76 1 1

0.68 0.91 0.74 0.71 0.78 1

0.81 1 0.73 0.73 0.89 0.91

.000 .000 .003 .019 .000 .000 .000

Abbreviations: PPV, positive predictive value; NPV, negative predictive value.

8

P. Morbini et al. Table 4

Disease evolution and patient survival Total (%)

Disease persistence Disease recurrence Death DFS (months) DSS (months)

9 (21.9) 5 (12.1) 13 (31.7) 14.2 (SD:11.5) 16.4 (SD:9.9)

HPV status by HR HPV mRNA ISH

Statistical significance

Positive (%)

Negative (%)

P

2 (10) 2 (10) 4 (20) 15.4 (SD:11.3) 16.6 (SD:10.3)

7 3 9 11.9 15.1

NS NS NS NS NS

(33,3) (14.2) (42.8) (SD:11.2) (SD:9.4)

Abbreviations: HR, high risk; HPV, human papilommavirus; ISH, in situ hybridization; DFS, disease-free survival; DSS, disease-specific survival; NS, nonsignificant.

amplified the conserved E6 gene were indeed more sensitive and correctly identified the presence of HPV16 in 6 of the 7 cases described above, while in the last case the positivity for HR DNA and mRNA probes and the negativity for all HPV16-specific tests, including mRNA ISH, suggest the presence of another fully integrated HR genotype. DNA ISH with the INFORM HPV-III probes showed in our study no false-positive results, and 10% false-negative rate. The good diagnostic performance of this highly automated assay was similar to what has been reported in most recent series [23]. False-negative results could be attributed to suboptimal sensitivity of DNA ISH assays (theoretically 1-2 viral copies per cell, but probably lower [32]) with respect to PCR, or to the presence of genotype variants that reduce probe sensitivity. Negative DNA ISH results in mRNA ISH-positive cases could also raise concern on the specificity of the second test. Non-specific binding of the mRNA probe to viral DNA has been recently described [25]; however, the finding of positive mRNA ISH in cases where no viral DNA was hybridized, and the difference in signal patterns frequently observed with the two hybridizations tests does not support the cross-hybridization hypothesis, while the dot-like nuclear signal sometimes observed with mRNA ISH has been suggested to represent the binding of the probe at “hot spots” of high viral mRNA transcription [33] rather than to genomic HPV DNA. We also noticed that prompt and brief (4-6 hours) formalin fixation resulted in more extensive cytoplasmic and nuclear stain with mRNA ISH, suggesting that signal distribution is a variable associated with the degree of mRNA degradation (data not shown). Considering cohort data, HR mRNA ISH-positive tumors were significantly more frequent in palatine and lingual tonsils than in other oropharyngeal subsites, and in non-smoking patients, and consistently showed a non-keratinizing histology, thus confirming most of the characteristics that have been so far associated with HPV-driven oncogenesis [11,13,17,23]. The favorable prognostic implications of transcriptionally active infection consistently reported by several authors [11,13,17,20,22,23,34,35] were not confirmed in our series, while p16 and DNA ISH positivity was significantly associated with better DSS. Highly heterogeneous treatment protocols, as well as short follow-up, probably account for the lack of strong prognostic correlation for our mRNA data.

5. Conclusions While no stand-alone molecular test was sufficiently accurate for classifying HPV-driven OPSCC [19], the high sensitivity and specificity of the established combination of p16 immunostain and DNA ISH, and their significant association with outcome even in the absence of HPV status-targeted treatments, suggests that the introduction of mRNA ISH, which is quite labor- and cost-intensive, is presently not required in the diagnostic laboratory routine. In situ assays provide the advantage of morphological characterization of tumor tissue, and avoid the risk of false-negative amplification results in samples with a low proportion of tumor cells; however PCR targeting the conserved regions of HPV16 and possibly other HR HPV genomes, associated with tissue microdissection, can be used to further classify p16-positive/DNA ISH-negative cases, ensuring that all patients evaluated for de-escalated protocol receive appropriate treatment.

Acknowledgments The research was partially supported by grants RC08017800/12 and RC08053903/12 from the Italian Health Ministry to the IRCCS Policlinico San Matteo Foundation, Pavia, G11961 from the Associazione Italiana per la Ricerca sul Cancro (AIRC), and Ricerca Finalizzata Ministero della Salute project code RF-2011-02351315.

References [1] Marur S, D'Souza G, Westra WH, Forastiere A. HPV-associated head and neck cancer: a virus-related cancer epidemic. Lancet Oncol 2010; 11:781-9. [2] Joseph AW, D'Souza G. Epidemiology of human papillomavirusrelated head and neck cancer. Otolaryngol Clin N Am 2012;45:139-64. [3] Jemal A, Simard EP, Dorell C, et al. Annual Report to the Nation on the Status of Cancer, 1975-2009, featuring the burden and trends in human papillomavirus (HPV)-associated cancers and HPV vaccination coverage levels. J Natl Cancer Inst 2013;105:175-201. [4] Ragin CCR, Taioli E. Survival of squamous cell carcinoma of the head and neck in relation to human papillomavirus infection: review and meta-analysis. Int J Cancer 2007;121:1813-20.

HPV in FFPE oropharyngeal carcinoma biopsies [5] Fakhry C, Westra WH, Li S, et al. Improved survival of patients with human papillomavirus-positive head and neck squamous cell carcinoma in a prospective clinical trial. J Natl Cancer Inst 2008;100:261-9. [6] Ang KK, Harris J, Wheeler R, et al. Human papillomavirus and survival of patients with oropharyngeal cancer. N Engl J Med 2010; 363:24-35. [7] Ang KK, Sturgis EM. Human papillomavirus as a marker of the natural history and response to therapy of head and neck squamous cell carcinoma. Semin Radiat Oncol 2012;22:128-42. [8] Lassen P. The role of human papillomavirus in head and neck cancer and the impact on radiotherapy outcome. Radiother Oncol 2010;95: 371-80. [9] Kimple RJ, Harari PM. Is radiation dose reduction the right answer for HPV-positive head and neck cancer? Oral Oncol 2014;50:560-4. [10] Mirghani H, Amen F, Moreau F, et al. Human papilloma virus testing in oropharyngeal squamous cell carcinoma: what the clinician should know. Oral Oncol 2014;50:1-9. [11] Jung AC, Briolat J, Millon R, et al. Biological and clinical relevance of transcriptionally active human papillomavirus (HPV) infection in oropharynx squamous cell carcinoma. Int J Cancer 2010;126:1882-94. [12] Rautava J, Syrjänen S. Biology of human papillomavirus infections in head and neck carcinogenesis. Head Neck Pathol 2012;6:S3-S15. [13] Weinberger PM, Yu Z, Haffty BG, et al. Molecular classification identifies a subset of human papillomavirus-associated oropharyngeal cancers with favourable prognosis. J Clin Oncol 2006;24:736-47. [14] Smeets SJ, Hesselink AT, Speel EM, et al. A novel algorithm for reliable detection of human papillomavirus in paraffin embedded head and neck cancer specimen. Int J Cancer 2007;121:2465-72. [15] Singhi AD, Westra WH. Comparison of human papillomavirus in situ hybridization and p16 immunohistochemistry in the detection of human papillomavirus-associated head and neck cancer based on a prospective clinical experience. Cancer 2010;116:2166-73. [16] Rietbergen MM, Leemans CR, Bloemena E, et al. Increasing prevalence rates of HPV attributable oropharyngeal squamous cell carcinomas in the Netherlands as assessed by a validated test algorithm. Int J Cancer 2013;132:1565-71. [17] Schache AG, Liloglou T, Risk JM, et al. Evaluation of human papilloma virus diagnostic testing in oropharyngeal squamous cell carcinoma: sensitivity, specificity and prognostic discrimination. Clin Cancer Res 2011;17:6262-71. [18] Schlecht NF, Brandwein-Gensler M, Nuovo GJ, et al. A comparison of clinically utilized human papillomavirus detection methods in head and neck cancer. Mod Pathol 2011;24:1295-305. [19] Jordan RC, Lingen MW, Perez-Ordonez B, et al. Validation of methods for oropharyngeal cancer HPV status determination in US cooperative group trials. Am J Surg Pathol 2012;36:945-54. [20] Ukpo OC, Flanagan JJ, Ma X, Luo Y, Thorstad WL, Lewis JS. Highrisk human papillomavirus E6/E7 mRNA detection by a novel in situ hybridization assay strongly correlates with p16 expression and patient outcomes in oropharyngeal squamous cell carcinoma. Am J Surg Pathol 2011;35:1343-50.

9 [21] Lewis Jr JS, Chernock RD, Ma XJ, et al. Partial p16 staining in oropharyngeal squamous cell carcinoma: extent and pattern correlate with human papillomavirus RNA status. Mod Pathol 2012;25:1212-20. [22] Gao G, Chernock RD, Gay HA, et al. A novel RT-PCR method for quantification of human papillomavirus transcripts in archived tissues and its application in oropharyngeal cancer prognosis. Int J Cancer 2013;132:882-90. [23] Schache AG, Liloglou T, Risk JM, et al. Validation of a novel diagnostic standard in HPV-positive oropharyngeal squamous cell carcinoma. Br J Cancer 2013;108:1332-9. [24] Tenti P, Romagnoli S, Silini E, et al. Analysis and clinical implications of K-ras gene mutations and infection with human papillomavirus types 16 and 18 in primary adenocarcinoma of the uterine cervix. Int J Cancer 1995;64:9-13. [25] Dreyer JH, Hauck F, Oliveira-Silva M, Barros MH, Niedobitek G. Detection of HPV infection in head and neck squamous cell carcinoma: a practical proposal. Virchows Arch 2013;462:381-9. [26] Bishop JA, Ma XJ, Wang H, et al. Detection of transcriptionally active high-risk HPV in patients with head and neck squamous cell carcinoma as visualized by a novel E6/E7 mRNA in situ hybridization method. Am J Surg Pathol 2012;36:1874-82. [27] Sartor M, Steingrimsdottir H, Elamin F, et al. Role of p16/MTS1, cyclin D1 and RB in primary oral cancer and oral cancer cell lines. Br J Cancer 1999;80:79-86. [28] Morbini P, Dal Bello B, Alberizzi P, et al. Oral HPV infection and persistence in patients with head and neck cancer. Oral Surg Oral Med Oral Pathol Oral Radiol 2013;116:474-84. [29] Duray A, Descamps G, Arafa M, et al. High incidence of high-risk HPV in benign and malignant lesions of the larynx. Int J Oncol 2011;39:51-9. [30] Walboomers JM, Jacobs MV, Manos MM, et al. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol 1999;189:12-9. [31] Guo M, Gong Y, Deavers M, et al. Evaluation of a commercialized in situ hybridization assay for detecting human papillomavirus DNA in tissue specimens from patients with cervical intraepithelial neoplasia and cervical carcinoma. J Clin Microbiol 2008;46:274-80. [32] Lizard G, Démares-Poulet MJ, Roignot P, Gambert P. In situ hybridization detection of single-copy human papillomavirus on isolated cells, using a catalyzed signal amplification system: GenPoint. Diagn Cytopathol 2001;24:112-6. [33] Lewis Jr JS, Ukpo OC, Ma XJ, et al. Transcriptionally-active high-risk human papillomavirus is rare in oral cavity and laryngeal/hypopharyngeal squamous cell carcinomas—a tissue microarray study utilizing E6/E7 mRNA in situ hybridization. Histopathology 2012;60:982-91. [34] Hoffmann M, Ihloff AS, Görögh T, et al. p16(INK4a) overexpression predicts translational active human papillomavirus infection in tonsillar cancer. Int J Cancer 2010;127:1595-602. [35] Holzinger D, Schmitt M, Dyckhoff G, Benner A, Pawlita M, Bosch FX. Viral RNA patterns and high viral load reliably define oropharynx carcinomas with active HPV16 involvement. Cancer Res 2012;72: 4993-5003.