Materials
We identified 236 consecutive LSCC samples from cases between 2000 and 2017 in SymPathy (Tieto, Finland). Slides were histopathologically re-examined (by KN and KO) for correct diagnosis and classification of cancer site.
Exclusion criteria for analysis were: 1) repeated biopsies from the same case (112 samples), 2) previous radiotherapy to the neck (39 samples), 3) classification errors (5 samples) and 4) too little tumor tissue for analysis (2 samples).
Patient characteristics
SymPathy data provided access to clinical chart databases to supplement patient characteristics.
Descriptive data presented in two groups
Descriptive data on 78 LSCC samples included in the PCR analysis of EBV, HCMV, HAdV and p16 (analyzed by IHC) are presented. Due to HPV DNA fragmentation, 45 LSCC samples could not be analyzed, and were therefore drop-outs in the HPV PCR results (Figure 1). To clarify the possible impact of the dropouts, we analyzed the dropout group separately in categories for age, sex, LSCC site (supraglottic, glottic, subglottic) and smoking status (non-smoker, former smoker, smoker) (Table 1).
Table 1. Case characteristics.
Results presented in numbers and proportion (in brackets). SD = standard deviation presented in square brackets.
Characteristics
|
Total group n =78 (%)
|
HPV subgroup n=33 (%)
|
Age
|
|
|
Mean [SD]
|
68.1 [±10.3]
|
67.2 [±10.2]
|
Range
|
41–86
|
46–83
|
Sex
|
|
|
Female
|
15 (19)
|
6 (18)
|
Male
|
63 (81)
|
27 (82)
|
Site
|
|
|
Supraglottic
|
12 (15)
|
5 (15)
|
Glottic
|
64 (82)
|
27 (82)
|
Subglottic
|
2 (3)
|
1 (3)
|
Smoking status
|
|
|
Non-smoker
|
14 (18)
|
6 (18)
|
Female
|
4 (5)
|
2 (6)
|
Male
|
10 (13)
|
4 (12)
|
Former smoker
|
21 (27)
|
8 (24)
|
Female
|
3 (4)
|
1 (3)
|
Male
|
18 (23)
|
7 (21)
|
Smoker
|
42 (54)
|
19 (58)
|
Female
|
7 (9)
|
3 (9)
|
Male
|
35 (45)
|
16 (49)
|
Smoking data missing
|
1 (1)
|
-
|
Cases
Seventy-eight (78) LSCC samples were assessed for eligibility. The case ages ranged from 41-86 with a mean of 68.1 years (± 10.3). The sex distribution was 15 (19%) women and 63 (81%) men, and other clinical characteristics are shown in Table 1.
HPV subgroup
Thirty-three LSCC samples underwent HPV genotyping, with an age range of 46-83, mean of 67.2 years (± 10.2). The sex distribution was 6 (18%) women and 27 (82%) men.
Methods
Combined PCR/microarrays (PapilloCheck®,Greiner BioOneGmbH, Frickenhausen, Germany) were used for detection and genotyping of HPV and real time-PCR for EBV, HCMV and HAdV DNA detection. EBER in situ hybridization (EBER-ISH) was used for EBV detection in slides from LSCC tumors. In addition, we used immunohistochemistry (IHC) for mapping of p16 (Figure 1).
PCR for HPV, EBV, HCMV and HAdV
DNA was prepared by cutting five 10‐μM sections from each LSCC biopsy followed by routine extraction using the QIAamp DNA FFPE Tissue Kit (QIAgen, Hilden, Germany). Quality and quantity of total DNA was measured spectrophotometrically.
HPV detection by PapilloCheck® requires a DNA fragment of 350 bp for successful PCR performance. To assess the degree of DNA fragmentation prior to detection and typing using the PapilloCheck®, we performed three different PCR assays amplifying amplicon lengths of 536, 268 or 100 bp of the ß-globin housekeeping genes, and one real-time PCR assay detecting 60 bp of the RNasP housekeeping gene.
The PapilloCheck® is a multiplex PCR method with fluorescent primers, for genotyping of 24 HPV-types including 18 high‐risk HPV types; HPV 16, 18, 31, 33, 35, 39, 45, 51, 52, 53, 56, 58, 59, 66, 68, 70, 73 and 82, and six low‐risk HPV types; HPV 6, 11, 40, 42, 43, and 44.
For EBV detection, a real-time PCR assay of the BamHI-W region of the EBV genome was used. The sequences of the forward and reverse primers were 5'-GCATAATGGCGGACCTAGG-3' and 5'-AAGATAGCAGCAGCGCAGC-3', respectively, flanking the dual-labelled probe 5'-(FAM)TAAAACCCCCAGGAAGCGGGTCTATG(TAMRA)-3´. Real-time PCR detection towards the polymerase gene of HCMV consisted of the amplification primers 5'- TGDGTCACGCCGTTCTCTAG - 3' and a 50:50 mix of 5' - ATCATCATGGCCCACAACCT- 3' and 5' - ATCATCATGGCTCACAACCT - 3', flanking the dual-labelled fluorescent probe 5'- (FAM)ACTCGCCACCCGGCACCAAC(TAMRA) - 3'.
Qualitative real-time PCR for the detection of HAdV was performed as described by Hernroth et al. [12].
EBER-ISH for EBV mapping
The EBER-ISH is an in-situ hybridization kit for EBER transcripts (Epstein-Barr Encoding Region), and this was applied for EBV mapping in tumor and connective tissue.
Two tissue sections were cut from each LSCC biopsy, one was incubated with an RNA-positive control probe (800-2846, Ventana). The second section was tested for EBER-1 and EBER-2 by the Inform EBER Probe (800-2842, Ventana Medical Systems, Roche Diagnostics GmbH, Mannheim Germany). NBT- BCIP Detection System (800-092, Roche Diagnostics GmbH, Mannheim, Germany) was employed for visualization. We used a Ventana BenchMark ULTRA system (Roche Diagnostics AG, Switzerland) for staining.
EBER-positive cells were counted in tumor and connective tissue by three of the authors (AS, AH and KN). Consensus was reached after comparison of individual results.
EBER was graded as positive when there were ≥ 2 positive cells and graded as negative when there were one or no positive cells.
p16 immunohistochemistry (IHC) and Quickscore analysis
To detect the p16 protein, five µm sections were cut from paraffin blocks. The staining system (Ventana BenchMark ULTRA, Roche Diagnostics AG, Switzerland) was applied, and slides pretreated in CC1 (cell condition solution 1, pH 8.4; Ventana) for 48 minutes. For detection of p16, CINtec® p16 Histology (Roche Diagnostics AG, Switzerland) was used. To visualize the antibody, the Opti View DAB IHC Detection Kit (Ventana Medical Systems) was employed.
p16 expression was evaluated by three members of the group (AS, AH and KN) using a light microscope (Nikon Eclipse Ci, Tokyo, Japan) and a semiquantitative Quickscore system by Detre et al. [13]. Using this system, we evaluated both proportion of p16 stained cells and the staining intensity (see Supplementary Information). Multiplying the scores for proportion of stained cells with intensity, results in a Quickscore ranging from 0–18. After comparison rounds consensus was reached.
Statistics
Descriptive statistics (IBM SPSS USA, version 26) were presented regarding age, gender, smoking, cancer site and presence of virus and p16. Cross tabulations were used to describe numbers and proportions of virus and p16. The association between virus and p16 were analyzed using Chi-square test and Fisher´s exact test. A probability value of equal or less than 0.05 (2-sided) was considered statistically significant.