Abstract
Available clinical human papilloma virus (HPV) diagnostics for head and neck cancer have limited sensitivity and/or fail to define the HPV genotype. Common HPV genotyping assays are costly and labor intensive. We sought to develop a next-generation sequencing (NGS)-based HPV genotyping assay that was sensitive enough to work on formalin-fixed paraffin-embedded (FFPE) samples. We developed an ion torrent NGS HPV genotyping assay using barcoded HPV PCR broad-spectrum general primers 5+/6+ (BSGP)5+/6+. To validate genotype specificity and use in archived clinical FFPE tumor samples, we compared NGS HPV genotyping at 2 sequencing centers with typing by Roche Linear Array assay in 42 oropharyngeal and cervical cancer specimens representing 10 HPV genotypes, as well as HPV-negative cases. To demonstrate the detection of a broad range of HPV genotypes, we genotyped a cohort of 266 cervical cancers. A comparison of NGS genotyping of FFPE cancer specimens with genotyping by Linear Array showed concordant results in 34/37 samples (92%) at sequencing site 1 and 39/42 samples (93%) at sequencing site 2. Concordance between sites was 92%. Designed for use with 10 ng genomic DNA, the assay detected HPV using as little as 1.25 ng FFPE-derived genomic DNA. In 266 cervical cancer specimens, the NGS assay identified 20 different HPV genotypes, including all 13 carcinogenic genotypes. This novel NGS assay provides a sensitive and specific high-throughput method to detect and genotype HPV in a range of clinical specimens derived from FFPE with low per-sample cost.
Keywords: human papillomavirus, viral, head and neck cancer
INTRODUCTION
Head and neck cancer is the sixth most common cancer worldwide.1 The majority of these tumors includes squamous cell carcinomas (SCCs) arising from various anatomic sites, including the oral cavity, oropharynx, larynx, and hypopharynx. Historically, head and neck SCC (HNSCC) has been associated with alcohol and tobacco use. More recently, it has become clear that HPV, long known to cause cervical and other genital cancers,2 also causes HNSCC, predominantly in the oropharynx.3, 4 HPV-associated oropharyngeal cancers, unrelated to alcohol or tobacco use, in many countries, are now the majority of cases,5 and are associated with substantially better clinical outcomes.6 Therefore, HPV has become an important biomarker in clinical decision making and prognostication of patients with oropharyngeal SCC.
As is the case with cervical cancers,7 HPV genotype 16 causes the majority of oropharyngeal cancers8; however, the incidence and influence of specific HPV genotypes in various populations are still being explored. In cervical cancer, 12 HPV genotypes are confirmed to be oncogenic (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, and 59), whereas several more are currently considered probably (68) and possibly (26, 53, 66, 67,70, 73, and 82) carcinogenic.9
Screening strategies for HPV-associated HNSCC depend on analysis of biopsied tumor material, whereas the vast majority of HPV detection performed in the clinical setting has been in the context of screening of cervical cytology specimens. Currently, there is no standard screening test for HPV in HNSCC, and testing varies among institutions. The most commonly used clinical tests are in situ hybridization to detect HPV DNA and p16INK4a (p16) immunostaining. Overexpression of the protein p16 in HPV-transformed tumor cells is strongly correlated with the presence of oncogenic HPV and improved outcomes in HNSCC,10–12 making it a useful surrogate marker for HPV-associated malignancy. In situ hybridization for HPV is highly specific for the specific genotypes tested (typically a pool of high-risk genotypes) but in studies in HNSCC, has frequently been found to lack sensitivity, resulting in unacceptable levels of false-negative results.10, 11, 13, 14 p16 immunostaining is sensitive but may be discordant with HPV status in a small number of cases.12, 15, 16 Neither method provides the specific HPV genotype.
In cervical cancer screening, 3 commercial platforms have been U.S. Food and Drug Administration approved for high-risk HPV detection (digene Hybrid Capture 2, Qiagen, Germantown, MD, USA; cobas, Roche Diagnostics, Indianapolis, IN, USA; and APTIMA, Hologic, Marlborough, MA, USA) in liquid cytology specimens only.17 None of these detection methods provides the specific HPV genotype.
To incorporate full HPV genotyping in studies of large cohorts, many investigators have relied on PCR-based strategies, where degenerate or pooled primer sets, capable of amplifying a significant range of HPV genotypes, are paired with hybridization to genotype-specific probes immobilized on beads, membrane arrays, or chips.12 The linear array HPV (LAHPV; Roche Diagnostics) genotyping test has commonly been used in this context12 and is effective in identifying 37 HPV genotypes (including all high-risk types) with high accuracy18 in both preserved cytology and FFPE samples. Unfortunately, the assay is costly and low throughput.
The advancement of NGS technology allows the development of a high-throughput, affordable assay for HPV genotyping. We developed an ion torrent NGS HPV genotyping assay using an established primer system with the capacity to detect the broadest range of HPV genotypes with minimal input DNA (≤10 ng), making this amenable for FFPE samples. To evaluate the ability of the new assay to genotype HPV accurately in archived FFPE clinical samples, we used it to genotype HNSCC and cervical carcinoma samples for direct comparison with genotyping by the LAHPV assay. To validate the new assay against a wide spectrum of HPV genotypes, 266 cervical cancers from a separate cohort were genotyped.
MATERIALS AND METHODS
Samples
Genomic DNA from 29 oropharyngeal HNSCC samples was prepared from FFPE tissue using the QIAamp DNA FFPE tissue kit for a previous study.19 In addition, DNA was extracted from FFPE tissue of 13 de-identified cervical cancer cases, also using the QIAamp DNA FFPE Tissue Kit. Following institutional review board (IRB) approval, all tumor samples were obtained from the University of Maryland Greenebaum Cancer Center (UMGCC) Pathology and Biorepository shared service.
The project was approved by the University of Maryland School of Medicine IRB. Where required, human subjects participating in the study have been given informed consent. All FFPE samples were obtained, following IRB approval from the UMGCC Pathology and Biorepository shared service. Additional cervical cancer specimens were provided according to the ethical approval cited.20
Two hundred sixty-six cervical cancer specimens from a separate cohort20 were provided in accordance with the ethical approvals cited. In brief, DNA was extracted from cervical cancer tissues (5–10 mg) and stored in RNAlater using the AllPrep DNA/RNA Micro Kit (Qiagen), as directed by the manufacturer.20
HPV16 PCR
PCR for the E6 and E7 genes of HPV16 was performed on the HNSCC cases for a previous study.19 DNA was extracted from several (3–5) 10 μ sections of FFPE oropharyngeal cancer tissue using the QIAamp DNA FFPE Tissue Kit (Qiagen), according to the manufacturer’s protocol. DNA was quantified using the Quant-iT dsDNA Assay Kit, High Sensitivity (Invitrogen, Thermo Fisher Scientific, Grand Island, NY, USA), and stored at −80°C in aliquots.
p16INK4a Immunohistochemistry
p16INK4a immunohistochemistry (IHC) was performed on these samples for a previous study.12 In brief, p16 IHC was performed using commercially available antibodies (clone JC8; Santa Cruz Biotechnologies, Dallas, TX, USA) and scored for cytoplasmic and nuclear staining by a consensus of 2 blinded pathologists. Only tumor cells with moderate or high intensity were counted. Proportional scoring was semiquantified as follows: 0, <10% staining; 1+, 10–49% staining; 2+, 50–70% staining; 3+, >70% staining. Scores of 2 or 3+ were defined as positive.
Roche Diagnostics LAHPV
The manufacturer’s protocol is designed to be performed starting with cervical cancer cells collected in preservative media and therefore, begins with DNA sample preparation instructions for that starting material. For our purposes, these steps were replaced by the DNA preparation method described above. The manufacturer’s protocol was followed beginning with the amplification step. For each sample, 100–500 ng DNA in 50 μl was included in the amplification reaction and tested along with the manufacturer’s positive and negative controls (Roche Diagnostics).21 The PCR machine specified by the protocol was used [96-well GeneAmp PCR System 9700 (Applied Biosystems, Thermo Fisher Scientific)]; however, the silver sample block used was not gold plated. For dilution series, the indicated quantity of DNA was used in the assay.
Ion Torrent HPV Genotyping Assay
The assay uses the BSGP5+/6+ primer system, which was designed by Schmitt et al.22 to amplify homogeneously a broad range of HPV genotypes and consists of a pool of 3 reverse primers and 9 forward primers. For use in the ion torrent system, the reverse primer sequences are modified by a preceding adapter sequence and 1 of 96 barcode sequences (Fig. 1), whereas the forward primer sequences are preceded by adapter sequence only. For each sample, a pool of 3 reverse primers with the same barcode is combined with the forward primer pool to amplify a library of barcoded HPV amplicons. The uniquely identified libraries are then pooled for emulsion PCR and sequencing.
Figure 1.
Schematic of ion torrent HPV genotyping assay. The HPV gene target in each DNA sample is amplified using fused primers based on the BSGP5+/6+ primer system. Reverse primers append an ion adapter and barcode before the gene-specific sequence. Forward primers append the ion adapter only.
The assay was tested initially at the U.S. National Cancer Institute-Frederick National Laboratory for Cancer Research (NCI-FNLCR). Investigators there were provided with the 29 HNSCC and 8 of the cervical carcinoma DNA samples completely blinded to LAHPV genotyping results. The identical ion torrent HPV genotyping assay was used in the Genomics Shared Service at UMGCC to analyze the same samples, along with 5 additional cervical cancer specimens. Genomic DNA (10 ng), as quantified by NanoDrop, was included in each HPV library amplification reaction. For dilution series, the indicated quantity was included in the reaction.
Library amplification reactions were analyzed using an Agilent BioAnalyzer for presence of product of the expected size (with adapter sequences, ∼150 bp). The investigators at NCI-FNLCR included only samples with an amplified product in the sequencing pool. The investigators at UMGCC included all samples in the sequencing pool at a standardized concentration of ∼500 pM, as determined by the BioAnalyzer. Samples without library product detection were included in the pool at equal volumes. Pooled samples were quantified for emulsion template preparation on the Qubit 2.0 Fluorometer and prepared using Ion PGM 200 kits on the OneTouch 2. Sequencing was performed on the ion torrent PGM using the 200 v2 sequencing chemistry and 316v2 chips.
Raw data collection and processing were performed by the ion torrent server, version 4.4.3, and mapped to the full genomic sequences of HPV downloaded from the PaVE database with a minimum quality score of AQ17. Further filtering of only reads >100 bp was performed using NGSUtils.23 A sample must contain >5000 reads in any HPV genotype to be called positive. If a coinfection of HPV is present, then the minor number of reads must total >1% of the total number of reads for that given sample. A sample with no reads and β-globin-positive amplification, using the primers BGMS3 (forward) AAT ATA TGT GTG CTT ATT TG and BGMS10 (reverse) AGA TTA GGG AAA GTA TTA GA, was called HPV negative.
RESULTS
Table 1 presents the results from two sequencing sites (NCI-FNLCR and UMGCC) of ion torrent NGS HPV genotyping of genomic DNA from FFPE tumors. All samples were genotyped by LAHPV for comparison. HPV16 DNA PCR and p16 immunostaining results, for most cases, were also available and are included for reference.
TABLE 1.
Comparison of HPV Genotyping of HNSCC and Cervical Cancer FFPE Samples by Roche LAHPV Genotyping Kit and by Ion Torrent Sequencing Assay
| Casea | LAHPVb |
Ion Torrent Seq NCI-FNLCRc |
Ion Torrent Seq UMGCCd |
HPV16-only PCR | p16INK4A IHC |
|---|---|---|---|---|---|
| Type | Type (no. of reads) | Type (no. of reads) | |||
| HN1 | HPV16 | HPV16 (91,932) | HPV16 (98,464) | Positive | Negative |
| HN2 | Negative | Negative | Negative | Negative | ND |
| HN3 | Negative | Negative | Negative | Negative | Negative |
| HN4 | Negative | Negative | Negative | Negative | Positive |
| HN5 | HPV16 | HPV16 (123,434) | HPV16 (331,951) | Positive | Positive |
| HN6 | HPV16 | HPV16 (135,707) | HPV16 (401,815) | Positive | Positive |
| HN7 | Negative | Negative | HPV33 (314,764) | Negative | Positive |
| HN8 | Negative | HPV16 (62,499) | Negative | Negative | Negative |
| HN9 | Negative | Negative | Negative | Negative | Negative |
| HN10 | HPV16 | HPV16 (98,430) | HPV16 (435,742) | Positive | Positive |
| HN11 | HPV16 | HPV16 (97,071) | HPV16 (475,595) | Positive | Positive |
| HN12 | HPV26 | HPV26 (64,317) | HPV26 (545,699) | Negative | Positive |
| HN13 | HPV35 | HPV35 (96,087) | HPV35 (386,719) | Negative | Positive |
| HN14 | Negative | Negative | Negative | Negative | Negative |
| HN15 | HPV16 | HPV16 (119,901) | HPV16 (198,461) | Positive | Positive |
| HN16 | Negative | Negative | Negative | Negative | Negative |
| HN17 | HPV16 | HPV16 (119,990) | HPV16 (156,232) | Positive | Positive |
| HN18 | Negative | HPV16 (80,702) | Negative | Negative | Negative |
| HN19 | HPV35 | HPV35 (84,022) | HPV35 (187,001) | Negative | Positive |
| HN20 | HPV16 | HPV16 (137,235) | HPV16 (124,467) | Positive | Positive |
| HN21 | HPV6 | HPV6 (101,171) | HPV6 (846,077) | Negative | Positive |
| HPV59 | HPV59 (3090) | HPV59 (657,207) | |||
| HN22 | HPV16 | HPV16 (116,149) | HPV16 (133,061) | Positive | Positive |
| HN23 | HPV16 | HPV16 (98,914) | HPV16 (121,619) | Positive | Positive |
| HN24 | HPV58 | HPV58 (72,968) | HPV58 (496,657) | Negative | Positive |
| HN25 | Negative | Negative | Negative | Negative | Negative |
| HN26 | Negative | Negative | Negative | Negative | Negative |
| HN27 | HPV33 | HPV33 (72,262) | HPV33 (176,115) | Negative | Positive |
| HN28 | Negative | Negative | Negative | Negative | Negative |
| HN29 | Negative | Negative | Negative | Negative | Negative |
| CC1 | HPV16 | HPV16 (89,642) | HPV16 (197,663) | ND | ND |
| CC2 | HPV58 | HPV58 (41,426) | HPV58 (103,980) | ND | ND |
| HPV16 (6915) | HPV16 (5175) | ||||
| CC3 | HPV69 | HPV69 (85,865) | HPV69 (105,000) | ND | ND |
| CC4 | HPV16 | HPV16 (89,029) | HPV16 (501,380) | ND | ND |
| CC5 | HPV16 | HPV16 (96,184) | HPV16 (198,444) | ND | ND |
| CC6 | HPV58 | HPV58 (92,864) | HPV58 (85,333) | ND | ND |
| CC7 | HPV16 | HPV16 (148,894) | HPV16 (260,259) | ND | ND |
| CC8 | HPV18 | HPV18 (111,461) | HPV18 (103,205) | ND | ND |
| CC9 | HPV58 | ND | HPV58 (57,243) | Negative | ND |
| CC10 | HPV16 | ND | HPV16 (77,968) | Positive | ND |
| CC11 | HPV45 | ND | HPV45 (90,287) | Negative | ND |
| CC12 | Negative | ND | HPV16 (66,333) | Negative | ND |
| CC13 | HPV16 | ND | HPV16 (57,328) | Positive | ND |
ND, not done.
HNSCC cases indicated by “HN” and cervical cancer cases by “CC.”
Result from LAHPV kit.
Blinded result from ion torrent sequencing performed at NCI-FNLCR.
Result from ion torrent sequencing performed at UMGCC.
The 29 HNSCC samples comprised both HPV-negative cases and cases positive for 7 different HPV genotypes (6, 16, 26, 33, 35, 58, and 59), according to LAHPV genotyping. Among the 13 cervical cancer cases, 5 HPV genotypes were represented (16, 18, 45, 58, and 69). Although all cervical cancers can be presumed to be HPV positive, LAHPV did not detect HPV in 1 case (CC12).
A comparison of the results of LAHPV with the UMGCC sequencing site showed concordant results in 28 of 29 HNSCC samples (97%) and 11 out of 13 cervical samples (85%). The discordant HNSCC sample (HN7) was p16 positive by IHC, HPV33 positive by ion torrent NGS genotyping, but negative by LAHPV. One discordant cervical case, CC12, was HPV16 positive by sequencing, where no HPV was detected by LAHPV. These 2 samples were repeated in triplicate by sequencing, and all 3 times yielded HPV33 positive (HN7) and HPV16 positive (CC12). The other discordant cervical case had HPV 58 (103,980 reads) and HPV16 (5175 reads) detected by NGS genotyping, whereas LAHPV detected only HPV 58. The sequencing of CC2 in triplicate yielded similar results, observing HPV58 and HPV16 sequences in this sample.
A comparison between the results of LAHPV with sequencing data from NCI-FNLCR showed strong concordance (34 out of 37). Sequencing of cervical sample CC2 by NCI-FNLCR revealed the same coinfected genotypes as described above, HPV58 and HPV16. Two cases (HN8 and HN18) were found to be HPV16 positive by NGS genotyping at NCI-FNLCR, whereas these same samples were HPV negative by LAHPV, HPV negative by NGS genotyping at UMGCC, p16 negative by IHC, and HPV16 negative by E6/E7 PCR. The DNA aliquots tested at NCI-FNLCR for HN8 and HN18 were subsequently tested in triplicate at UMGCC and found to be negative for HPV16 in all 3 replicates. These results lead us to believe that the positive results observed at NCI-FNLCR for HN8 and HN18 represent contaminants introduced into the sequencing reaction from a source other than the sample aliquots provided and do not represent a false-positive result.
With the comparison of the sequencing results at the UMGCC and NCI-FNLCR labs, again, 34 of 37 cases (92%) were concordant. This included the 2 contaminants identified above (HN8 and HN18) and 1 sample, HN7, which was determined to be HPV negative at NCI-FNLCR and HPV33 positive at UMGCC. This sample failed to produce an ∼150 bp band at NCI-FNLCR and therefore, was not included in the sequencing reaction. However, UMGCC observed a faint ∼150 bp for this sample and upon sequencing, revealed an HPV 33 genotype. The repetition of this sample, using the DNA aliquot from NCI-FNLCR, confirmed the HPV33 genotype for this sample.
Twenty-eight of 29 HNSCC samples had data available for p16 overexpression. Concordance between HPV detection and p16 expression was 93% (26/28) when HPV was detected by NGS genotyping at UMGCC, 89% (25/28) by LAHPV, and 82% (23/28) by NGS genotyping at NCI-FNLCR. In case HN1, all DNA testing methods, including PCR for HPV16 E6/E7, detected HPV16, whereas p16 was negative, indicating that this tumor is likely a true HPV/p16 discordant case, where p16 expression has been lost.
Both the sequencing assay and LAHPV were highly sensitive in our hands. A known positive case was used to explore the sensitivity of both assays. In a serial dilution of starting DNA (10, 5, 2.5, 1.25, and 0.625 ng), we were able to detect HPV16 by ion torrent NGS genotyping with as little as 1.25 ng DNA (obtaining 263,813 reads). The detection limit for this sample in the LAHPV assay was 2.5 ng DNA.
A large cohort of cervical cancer cases was tested by ion torrent NGS genotyping at NCI-FNLCR20; the genotypes detected were not reported previously. Twenty HPV genotypes were detected in this cohort, including all 13 high-risk genotypes (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68). Types 6, 26, 44, 53, 67, 69, and 81 were also detected.
DISCUSSION
There is currently a need for a sensitive, specific HPV genotyping assay that is rapid, cost effective, and appropriate for large sample sets. Such an assay could have great use for population studies and for routine diagnosis of HPV-associated cancers in a clinical setting. Current HPV vaccines protect against a limited number of HPV genotypes. The most common HPV genotypes in oropharyngeal cancer are 16, 18, 33, 35, and 58,24 and in our cohort of 185 oropharyngeal cancers, genotype 35 was the most common non-16 genotype (2 cases).12 It is noteworthy that HPV 35 is not among the genotypes protected against by the 9-valent Gardasil vaccine. Description and detection of HPV genotypes, subtypes, and variants, prevalent in different cancer types and within different regions and populations worldwide, are ongoing endeavors, better served by sequencing, where the technique allows identification of previously unknown sequence variation rather than by hybridization-based techniques that fail to provide a readout of the actual sequence present.
Studies examining the HPV genotype in HNSCC have commonly relied on post-PCR hybridization methods, such as line-blot assays.12 Whereas the LAHPV test is somewhat sensitive and specific, it is very costly on a per-sample basis, as well as labor intensive, making it impractical for many large studies. It has the advantage of comparatively modest up-front equipment costs, but it is not approved for routine diagnostic use in the United States.
Any HPV detection and/or genotyping method rely on the ability to amplify all relevant HPV genotypes. Our test is based on a pool of primers, BSGP5+/6+, which compared with the formerly standard GP5+/6+ PCR, was shown to match or significantly improve amplification of 24 HPV genotypes, displays improved primer alignment to 50 types and in a cohort of 1085 clinical samples, detected 26 HPV genotypes.21 The LAHPV test is based on the PGMY 09/11 primer pool and is reported by the manufacturer to detect 37 HPV genotypes. The significant agreement between our test and LAHPV, as well as the detection of 20 genotypes in a set of cervical cancer specimens, suggests that the primer system that we selected is fully capable of robust amplification of a broad spectrum of HPV genotypes.
The NGS assay presented here is robust, returning excellent read numbers from FFPE samples. It is highly sensitive, specific, and capable of identifying multiple coinfected HPV genotypes in the same sample. Identification of coinfection by HPV16 by sequence methods and identification of an HPV16-positive cervical cancer sample, all of which were not identified by LAHPV, demonstrate the increased sensitivity of NGS sequencing. These data also confirm the ability of LAHPV to give false-negative results. Most importantly, the assay is reproducible across laboratories. Of the discordant samples identified between the LAHPV method and NGS sequencing at 2 independent sites, all but 1 sample can be explained as a contaminant (2 samples sequenced at NCI-FNLCR) or as a result of the increased sensitivity of the NGS sequencing approach. When we account for the discrepancies between the different platforms and sites, there was essentially a near-perfect concordance across the data. Additionally, the data correlate well with direct genotype-specific PCR and p16 IHC assays. The barcoding of the primer sets would permit larger runs of 96 to 384 samples simultaneously with rapid turnaround time and modest technician time. Whereas the NGS assay has a significant capital equipment cost, the per-sample reagent cost is less than 1/10 of the cost associated with the LAHPV test. As such, a single laboratory could readily support multiple epidemiologic studies or a number of clinical institutions. This study demonstrates the ability of this NGS sequencing assay as an extremely sensitive and accurate method to genotype HPV. Ideally, this could become an assay of choice for diagnostic laboratories; however, additional steps would need to be taken to perform a clinical validation following standards for a laboratory-developed test (including the establishment of accuracy, specificity, precision, and limit of detection).
ACKNOWLEDGMENTS
The authors acknowledge Enrique Alvirez, Eduardo Gharzouzi, and Michael Dean, who provided the cervical cancer specimens used to validate HPV genotype sensitivity. This work was funded by the U.S. National Institutes of Health National Cancer Institute Grant P30-CA134274; and Orokawa Foundation.
REFERENCES
- 1.Ferlay J, Shin HR, Bray F, Forman D, Mathers C, Parkin DM. Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer 2010;127:2893–2917. [DOI] [PubMed] [Google Scholar]
- 2.zur Hausen H. Papillomaviruses and cancer: from basic studies to clinical application. Nat Rev Cancer 2002;2:342–350. [DOI] [PubMed] [Google Scholar]
- 3.D’Souza G, Kreimer AR, Viscidi R, Pawlita M, Fakhry C, Koch WM, Westra WH, Gillison ML. Case-control study of human papillomavirus and oropharyngeal cancer. N Engl J Med 2007;356:1944–1956. [DOI] [PubMed] [Google Scholar]
- 4.Gillison ML, Koch WM, Capone RB, Spafford M, Westra WH, Wu L, Zahurak ML, Daniel RW, Viglione M, Symer DE, Shah KV, Sidransky D. Evidence for a causal association between human papillomavirus and a subset of head and neck cancers. J Natl Cancer Inst 2000;92:709–720. [DOI] [PubMed] [Google Scholar]
- 5.Chaturvedi AK, Anderson WF, Lortet-Tieulent J, Curado MP, Ferlay J, Franceschi S, Rosenberg PS, Bray F, Gillison ML. Worldwide trends in incidence rates for oral cavity and oropharyngeal cancers. J Clin Oncol 2013;31:4550–4559. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ang KK, Harris J, Wheeler R, Weber R, Rosenthal DI, Nguyen-Tân PF, Westra WH, Chung CH, Jordan RC, Lu C, Kim H, Axelrod R, Silverman CC, Redmond KP, Gillison ML. Human papillomavirus and survival of patients with oropharyngeal cancer. N Engl J Med 2010;363:24–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Monsonego J, Cox JT, Behrens C, Sandri M, Franco EL, Yap PS, Huh W. Prevalence of high-risk human papilloma virus genotypes and associated risk of cervical precancerous lesions in a large U.S. screening population: data from the ATHENA trial. Gynecol Oncol 2015;137:47–54. [DOI] [PubMed] [Google Scholar]
- 8.Kreimer AR, Clifford GM, Boyle P, Franceschi S. Human papillomavirus types in head and neck squamous cell carcinomas worldwide: a systematic review. Cancer Epidemiol Biomarkers Prev 2005;14:467–475. [DOI] [PubMed] [Google Scholar]
- 9.IARC Working Group on the Evaluation of Carcinogenic Risks to Humans Human papillomaviruses. IARC Monogr Eval Carcinog Risks Hum 2012;100B:255–313. [Google Scholar]
- 10.Rischin D, Young RJ, Fisher R, Fox SB, Le QT, Peters LJ, Solomon B, Choi J, O’Sullivan B, Kenny LM, McArthur GA. Prognostic significance of p16INK4A and human papillomavirus in patients with oropharyngeal cancer treated on TROG 02.02 phase III trial. J Clin Oncol 2010;28:4142–4148. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Mellin Dahlstrand H, Lindquist D, Björnestål L, Ohlsson A, Dalianis T, Munck-Wikland E, Elmberger G. P16(INK4a) correlates to human papillomavirus presence, response to radiotherapy and clinical outcome in tonsillar carcinoma. Anticancer Res 2005;25:4375–4383. [PubMed] [Google Scholar]
- 12.Liu SZ, Zandberg DP, Schumaker LM, Papadimitriou JC, Cullen KJ. Correlation of p16 expression and HPV type with survival in oropharyngeal squamous cell cancer. Oral Oncol 2015;51:862–869. [DOI] [PubMed] [Google Scholar]
- 13.Schlecht NF, Brandwein-Gensler M, Nuovo GJ, Li M, Dunne A, Kawachi N, Smith RV, Burk RD, Prystowsky MB. A comparison of clinically utilized human papillomavirus detection methods in head and neck cancer. Mod Pathol 2011;24:1295–1305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Chernock RD, Zhang Q, El-Mofty SK, Thorstad WL, Lewis JS Jr. Human papillomavirus-related squamous cell carcinoma of the oropharynx: a comparative study in whites and African Americans. Arch Otolaryngol Head Neck Surg 2011;137:163–169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Jordan RC, Lingen MW, Perez-Ordonez B, He X, Pickard R, Koluder M, Jiang B, Wakely P, Xiao W, Gillison MI. Validation of methods for oropharyngeal cancer HPV status determination in US cooperative group trials. Am J Surg Pathol 2012;36:945–954. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Westra WH. Detection of human papillomavirus (HPV) in clinical samples: evolving methods and strategies for the accurate determination of HPV status of head and neck carcinomas. Oral Oncol 2014;50:771–779. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Binnicker MJ, Pritt BS, Duresko BJ, Espy MJ, Grys TE, Zarka MA, Kerr SE, Henry MR. Comparative evaluation of three commercial systems for detection of high-risk human papillomavirus in cervical and vaginal ThinPrep PreservCyt samples and correlation with biopsy results. J Clin Microbiol 2014;52:3763–3768. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Black CC, Bentley HA, Davis TH, Tsongalis GJ. Use of a linear array for the detection of human papillomavirus genotypes in head and neck cancer. Arch Pathol Lab Med 2010;134:1813–1817. [DOI] [PubMed] [Google Scholar]
- 19.Zandberg DP, Liu S, Goloubeva OG, Schumaker LM, Cullen KJ. Emergence of HPV16-positive oropharyngeal cancer in black patients over time: University of Maryland 1992–2007. Cancer Prev Res (Phila) 2015;8:12–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Lou H, Villagran G, Boland JF, Im KM, Polo S, Zhou W, Odey U, Juárez-Torres E, Medina-Martínez I, Roman-Basaure E, Mitchell J, Roberson D, Sawitzke J, Garland L, Rodríguez-Herrera M, Wells D, Troyer J, Castillo Pinto F, Bass S, Zhang X, Castillo M, Gold B, Morales H, Yeager M, Berumen J, Alvirez E, Gharzouzi E, Dean M. Genome analysis of Latin American cervical cancer: frequent activation of the PIK3CA pathway. Clin Cancer Res 2015;21:5360–5370. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Canadian Women’s HIV Study Group Enhanced detection and typing of human papillomavirus (HPV) DNA in anogenital samples with PGMY primers and the linear array HPV genotyping test. J Clin Microbiol 2006;44:1998–2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Schmitt M, Dondog B, Waterboer T, Pawlita M. Homogeneous amplification of genital human alpha papillomaviruses by PCR using novel broad-spectrum GP5+ and GP6+ primers. J Clin Microbiol 2008;46:1050–1059. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Breese MR, Liu Y. NGSUtils: a software suite for analyzing and manipulating next-generation sequencing datasets. Bioinformatics 2013;29:494–496. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Ndiaye C, Mena M, Alemany L, Arbyn M, Castellsagué X, Laporte L, Bosch FX, de Sanjosé S, Trottier H. HPV DNA, E6/E7 mRNA, and p16INK4a detection in head and neck cancers: a systematic review and meta-analysis. Lancet Oncol 2014;15:1319–1331. [DOI] [PubMed] [Google Scholar]