Aim, design and setting of the study
Response to cetuximab therapy in patients with HNSCC is only around 20%, and currently, neither prognostic nor predictive biomarkers have been identified to stratify patients with respect to potential response vs. non-response to cetuximab.
The aim of the present study was to investigate the HPV status, and in particular the Aurora kinase A polymorphism, in terms of their predictive and prognostic value in patients with HNSCC treated with cetuximab therapy.
Clinical data from patients with HNSCC was retrospectively analyzed. In vitro experiments were performed to complement clinical data collected.
Clinical data and experimental procedures
Patient population
FFPE tissues from 434 patients with squamous cell carcinoma of the oral cavity, oropharynx, hypopharynx, and larynx were used. Both tumor tissue and corresponding normal tissue from the same patient were examined. The included patients were treated at the ENT hospital, Klinikum rechts der Isar, Technical University Munich, at the ENT department of the University of Regensburg and the ENT department of the University of Heidelberg. The patients received either surgery, primary chemoradiotherapy (combined with cetuximab in many cases), or palliative chemo- or immunotherapy. The inclusion criterion was admittance to one of the three hospitals from 1988 to 2015. Histological and clinical data were collected from medical records, and information from the Munich Cancer Registry was used for the collection of survival data. Patient characteristics are enumerated in Table 1. All tumors were reclassified according to the International Union Against Cancer (UICC) TNM Classification System of 2002. The study was approved by the Ethics Committee of the Klinikum rechts der Isar, Technical University Munich (project number 1420/05).
Table 1
Patient characteristics | HNSCC (n = 434) |
---|
Age – years |
Median | 73 | |
Range | 56–90 | |
Sex |
Male | 362 | 83.4% |
Female | 72 | 16.6% |
T-Stage |
T1 | 42 | 9.7% |
T2 | 114 | 26.3% |
T3 | 100 | 23.0% |
T4a | 124 | 28.6% |
T4b | 1 | 0.2% |
Tx | 1 | 0.2% |
Unknown | 52 | 12.0% |
N-Stage |
N0 | 132 | 30.4% |
N1 | 41 | 9.4% |
N2a | 9 | 2.1% |
N2b | 93 | 21.4% |
N2c | 92 | 21.2% |
N3 | 10 | 2.3% |
Nx | 3 | 0.7% |
Unknown | 54 | 12.4% |
M-Stage |
M0 | 348 | 80.2% |
M1 | 13 | 3.0% |
Mx | 20 | 4.6% |
Unknown | 53 | 12.2% |
Grading |
G1 | 21 | 4.8% |
G2 | 185 | 42.6% |
G3 | 130 | 30.0% |
Gx | 7 | 1.6% |
Unknown | 91 | 21.0% |
Stage |
Stage I | 29 | 6.7% |
Stage II | 55 | 12.7% |
Stage III | 58 | 13.4% |
Stage IVA | 213 | 49.1% |
Stage IVB | 17 | 3.9% |
Stage IVC | 6 | 1.4% |
Unknown | 51 | 11.8% |
Localisation |
Oral cavity | 85 | 19.6% |
Oropharynx | 151 | 34.8% |
Hypopharynx | 94 | 21.7% |
Larynx | 99 | 22.8% |
Other | 4 | 0.9% |
Alcohol consumption |
Regularly | 183 | 42.2% |
Irregular | 80 | 18.4% |
Unknown | 171 | 39.4% |
Tobacco use |
Smoker | 208 | 47.9% |
Non-/ Ex-smoker | 80 | 18.4% |
Unknown | 146 | 33.6% |
PCR-restriction fragment length polymorphism (RFLP)
DNA was isolated from cell lines and FFPE tumor samples using the DNeasy Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions.
Specific PCR for AurkA genotypes at the Phe31Ile site was performed using the Taq-DNA Polymerase All-inclusive Kit (PeqLab, Erlangen, Germany) according the manufacturer’s guidelines. We applied the following PCR program: 95 °C for 5 min, 30 cycles of 95 °C for 30 s, 60 °C annealing for 30 s, 72 °C for 30 s, and 72 °C for 7 min on a thermocycler (BioRad, Munich, Germany). The following primers were used: AurkA forward: CTTTCATGAATGCCAGAAAGTT; AurkA reverse: CTGGGAAGAATTTGAAGGACA. The 165-bp PCR products were digested with ApoI (New England BioLabs, Inc., Beverly, USA) and separated on a 2.5% agarose gel. Digestion of the AurkA 31Phe allele results in two fragments (153 bp and 12 bp); since the AurkA 31Ile allele contains an additional third ApoI restriction site, three fragments (89 bp, 64 bp, and 12 bp) are generated after digestion. The results were further confirmed by DNA sequencing using an ABI 3100 DNA sequencer (Life Technologies, Darmstadt, Germany).
Immunohistochemical study
Fresh 1.5 µm sections were transferred to glass slides, deparaffined, and rehydrated. An antigen retrieval method (microwave activation in citrate buffered saline) was applied following the instructions provided by the manufacturer. After cooling, the slides were incubated with the following antibodies: Aurora-Kinase A (rabbit, clone 1F8) 1:200 (Cell Signaling Technology, Frankfurt, Germany), Aurora-Kinase B (rabbit) 1:200 (Cell Signaling Technology), Survivin (rabbit, clone 12C4) 1:100 (Dako Agilent Technologies, Hamburg, Germany), p-Akt Ser473 (rabbit, clone 736E11) 1:20 (Cell Signaling Technology). The reaction was developed with the labeled streptavidin-biotin-peroxidase system. The expression of p16INK4a was assessed using an auto staining system (Leica Bond-Max), the BOND Polymer Refine Detection Kit (both from Leica Microsystems GmbH), and a monoclonal mouse antibody (Roche Diagnostics GmbH, Mannheim, Germany). 3,3'-Diaminobenzidin (DAB) was used as the reaction indicator. After counterstaining with hematoxylin, slides were dehydrated in ascending concentrations with ethanol and mounted. Tissues with known expression of the respective antigen were used as positive controls, and for negative controls, irrelevant antibodies with the correct immunoglobulin isotypes were used (Fig. 5).
Scoring system for protein expression in immunohistochemical staining
A scoring system was applied to describe the expression levels of the proteins AurkA, AurkB, Survivin, and p-Akt Ser473. The stained tumor areas were dichotomized as follows: Adopted from Schauer et al., we used an immunostaining score comprised of intensity and a stained tumor area that had values between 0 and 7 [40]. To perform the statistical analysis, we set a cut-off at 2 and divided the samples into positive and negative. p16INK4a was considered to be positive when it was defined as strong and diffuse nuclear and cytoplasmic staining in ≥ 70% of the tumor cells, which is the same scoring criteria used in the study by Ang et al. [41].
Cell culture
For the investigation of the influence of the Aurora kinase A polymorphism and the HPV status in vitro, four suitable cell lines were used for the proliferation analyzes: UD-SCC-2 (AurkA Phe/Phe, HPV positive), UD-SCC-5 (AurkA Phe/Phe, HPV negative), UP-SCC-154 (AurkA Phe/Ile, HPV positive), SAS (AurkA Phe/Ile, HPV negative).
The cell lines UP-SCC-154 and SAS were obtained from DSMZ (Braunschweig, Germany), and the cell lines UD-SCC-2 and UD-SCC-5 were provided by the University of Düsseldorf (Clinic for Otolaryngology, Düsseldorf, Germany).
Cells were cultured in Dulbecco's modified Eagle medium (DMEM) or RPMI 1640 medium (Invitrogen, Darmstadt, Germany) containing 10% fetal calf serum (FCS) (Biochrom, Berlin, Germany), 2 mM glutamine, 100 µg/ml streptomycin, and 100 U/ml penicillin (Biochrom) and maintained at 37 °C in an atmosphere of 5% CO2 grown to a 70–90% confluence. Cells were treated with docetaxel (0.25 nM) (Selleckchem, Houston, TX, USA) and cetuximab (Erbitux®) (50 nM) (Merck KGaA, Darmstadt, Germany).
Analysis of cell proliferation
The binding of crystal violet to cellular DNA was used to assess cell proliferation via photometry. Cells were seeded at a density of 5 × 103 cells per well in six-well plates 24 h before treatment. Ten days after treatment with the inhibitors, the culture medium was aspirated and 500 µL of 4% formaldehyde was added to each well for 15 minutes. After washing with 0.1% Triton-X100/1x PBS and H2O, crystal violet (0.04%) was added to the fixed cells and incubated for 30 minutes. Finally, after removing excess crystal violet, SDS (1%) was added, and the optical density was measured at 595 nm using a microplate photometer after 1 h.
Colony formation assay
Cell survival after treatment with inhibitors was assessed using a colony formation assay (CFA). The cells were seeded in six-well plates (5 × 102 cells per well) and allowed to adhere overnight at 37 °C. The following day, the cells were treated and incubated at 37 °C for 10 days. The cell colonies were then formalin fixed (4% formaldehyde) and visualized by crystal violet (0.04%) staining (Sigma-Aldrich, Steinheim, Germany). The colonies were counted after rinsing off the dye. Colony numbers were depicted as the percentage of colonies from untreated cells.
Western blot
For protein expression analysis, cells were lysed in 1x lysis buffer (New England Biolabs, Frankfurt, Germany) supplemented with 1 mM PMSF (Roth, Karlsruhe, Germany). Equal amounts of protein (15 µg) were separated by SDS-PAGE and transferred to Immobilion membranes (Millipore, Schwalbach, Germany). Blocking of unspecific binding sites was performed using 5% (w/v) non-fat dry milk powder or 5% bovine serum albumin in 1x TBST. Membranes were incubated with primary antibodies diluted in 1x TBST for 14–16 hours at 4 °C. HRP-conjugated immunoglobulins (diluted 1:5000 in 5% non-fat dry milk/1x TBST) served as detection antibodies and were probed for 1 h at room temperature. Immunoreactivity was visualized using ChemiDocTM XRS + with ImageLab software 6.0 (BioRad, Munich, Germany). We used primary antibodies against p-Akt Ser473 (rabbit, clone D9E) 1:500 (Cell Signaling), Survivin (rabbit) 1:1000 (Cell Signaling), Aurora-Kinase A (rabbit) 1:500, Aurora-Kinase B (rabbit) 1:500, and Tubulin (rabbit) 1:5000.
Aneuploidy analysis by flow cytometry
After seeding in culture flasks (75 cm2, 5 × 105 cells per flask) UD-SCC-5 and SAS cells were fixed/permeabilized with 70% methanol overnight. Finally, cells were washed twice with 1x PBS and resuspended in 1x PBS containing 1 µg/ml DAPI 15 min prior to analysis. 1 × 105 DAPI stained cells of every sample were analyzed using a BD FACSCanto-II™ flow cytometer (BD Biosciences, San Jose, USA). DNA histograms were plotted on a linear scale and cell cycle fractions, i.e., percentages of cells in G0/G1-, S-, and G2/M-phase, were quantified using ModFit LT 3.2 software (Verity Software House, Topsham, ME, USA) upon cell doublet, aggregate, and debris discrimination via pulse processing.
Statistical analysis
Statistical analysis was performed using GraphPad Prism software, version 7.0e. Assuming a symmetry correlation structure for all experiments, all hypotheses were tested with one-way ANOVA test, and the means of treated cells and untreated controls were compared by way of a t-test. The minimum level of statistical significance was set at p < 0.05.
SPSS software version 23 was used to evaluate the clinical data. Using Kaplan-Meier analysis and the log rank test, the prognostic value of individual markers was evaluated. Correlation analysis for the clinical data (T-classification, lymph node metastasis, distant metastasis) was performed using a cross tabulation/fourfold table (chi-square test, Fisher’s exact test).