Evaluation of SHP1-P2 methylation as a biomarker of lymph node metastasis in patients with squamous cell carcinoma of the head and neck

The present study was approved by the Institutional Review Board (IRB) of the Faculty of Medicine, Chulalongkorn University, Thailand (IRB No. 516/56, approval No. 004/2014). All samples were coded-anonymized formalin-fixed, paraffin-embedded tissue retrieved retrospectively from the Department of Pathology, Faculty of Medicine, Chulalongkorn University, affiliated with King Chulalongkorn Memorial Hospital, a tertiary care, university teaching hospital in central Bangkok, Thailand. The samples included 10 normal tonsils, and samples from 51 patients diagnosed with HNSCC including 9 samples of primary HNSCC and 42 lymph nodes from patients with and without HNSCC metastasis as inclusion criteria. With the exception of one 11-year-old boy, all other patients were in their fourth decade or older at the time of clinical sampling. Lymph nodes had been derived from radical neck dissections and were classified into 3 types: (1) lymph nodes from patients with control cases of nonmetastatic HNSCC (LN N0, n = 15), (2) lymph nodes from patients with cases of metastatic HNSCC, but histologically negative for carcinoma cells (LN– Nx, n = 18), and (3) lymph nodes from patients with matched cases of metastatic HNSCC that were histologically positive for carcinoma cells (LN+ Nx, n = 18). The normal tonsils and the LN N0 and LN– Nx types were confirmed histologically to be free of carcinoma cells, whereas lymph nodes of the LN+ Nx type were diagnosed as having metastatic malignant cells in at least 70% of the nodal tissue. All samples were confirmed as meeting the criteria by consensus from 2 pathologists (NK and SK). The demographic data of the patients, clinical stages, and histological grades were reviewed from their medical records and are presented in Table 1.

All formalin-fixed, paraffin-embedded (FFPE) tissues were sliced into 3 to 5 sections of 5 µm thickness and left unstained. Another section was stained with hematoxylin and eosin (HE) for histopathological confirmation. In the HNSCC groups, tumor cells were manually microdissected as previously reported [13]. In brief, FFPE tissues were cut serially into 5 levels. The first and the last of the 5 slides were stained with HE. Selected areas on the first slide were circled using a marker pen. Subsequently selected areas on the last slide were also marked using the first slide as a reference and subsequently examined under a microscope. If the last slide was correctly marked, the remaining unstained slides (levels 2–4) would be processed in the same manner using the first and last HE-stained slides as references to select the areas. Later, the marked areas on the unstained slides were dissected using a standard 21-gauge needle. After deparaffinization with xylene, the DNA was isolated using Tris-buffered sodium dodecyl sulfate with proteinase K and then left at 50°C overnight, followed by phenol–chloroform extraction and ethanol precipitation. The isolated genomic DNA was eluted and then used for bisulfite treatment. Bisulfite modification of the genomic DNA was performed using an EZ DNA Methylation Kit (Zymo Research). The bisulfite-treated DNA was resuspended in 20 µL of water and stored at –20°C until needed for use.

Quantification of SHP1-P2 methylation levels was performed using a combination of methylation-specific primers and the absolute quantitative TaqMan probe real-time PCR technique, as previously published [12]. Briefly, the methylated and unmethylated sets were used to quantitate the amount of methylated and unmethylated DNA, respectively. The methylation set contained 5ʹ-AGGATTTATT CGATGATAGTTGTTATCGTT-3ʹ as the forward primer, 5ʹ-CTCCACCAACTACTTTTACGCAAC-3ʹ as the reverse primer, and 5ʹ-FAM-CTAACCCACGCTAATAA-MGB-3ʹ as the probe. The unmethylated set comprised 5ʹ-GTAGGATT TATTTGATGATAGTTGTTATTG-3ʹ as the forward primer, 5ʹ-TCCTCCACCAACTACTTTTACACAA-3ʹ as the reverse primer, and 5ʹ-VIC-CCCTAACCCACACTA-MGB-3ʹ as the probe. The 2 PCR sets were normalized with the β-actin control set, which was composed of the forward primer 5ʹ-GTGTATTTGATTTTTGAGGAGA-3ʹ, the reverse primer 5ʹ-CCTTAATACCAACCTACCCAA-3ʹ, and the probe 5ʹ-Cy5-AAGGTGAAYGTGGATGAAGTTGGTGGTGAGG-BHQ-3ʹ.

All PCRs were performed as duplex PCR using 2× TaqMan GTXpress Real-Time PCR Master Mix (Applied Biosystems) with an ABI 7500 fast real-time PCR system. The SHP1-P2 methylation level was calculated using the following equation:

where nM (normalized methylated DNA) and nU (normalized unmethylated DNA) are the levels of methylated DNA or unmethylated DNA, respectively, divided by the level of β-actin in the same reaction. The copy numbers for methylated DNA, unmethylated DNA, and β-actin were obtained by comparing the change in the fluorescence signal (ΔRN) of the target DNA minus the RN from a passive reference dye for a given reaction of each target with a standard curve generated using varying concentrations of each target. To determine whether SHP1-P2 methylation is epithelial carcinoma cell specific, tonsil tissue and microdissected HNSCC epithelia were used as negative and positive controls, respectively.

All statistical analyses were conducted using IBM SPSS Statistics for Windows (version 22). Descriptive data for the SHP1-P2 methylation levels are expressed as mean ± standard deviation. Independent sample Student t tests were conducted to determine the differences between all groups except for LN– Nx and LN+ Nx, for which a paired Student t test was applied. P< 0.05 was considered as significant. A receiver operating characteristic curve analysis was conducted to assess the ability of the SHP1-P2 methylation level to differentiate between lymph nodes with or without metastases.

The level of SHP1-P2 methylation from the tonsil samples was 10.27 ± 4.05% (n = 10) compared with 61.31 ± 17.00% (n = 9) in primary microdissected HNSCC epithelia (Figure 1A).

Figure 1

The SHP1-P2 methylation in LN N0, LN– Nx, and LN+ Nx was 10.27 ± 4.05%, 14.49 ± 10.03%, and 41.01 ± 24.51%, respectively. A significant difference was observed between LN N0 and LN+ Nx. In addition, the paired t test showed a significant difference (P< 0.0001) between LN– Nx and LN+ Nx (Figure 1Aand B and Table 2). There were no significant differences in SHP1-P2 methylation levels between different sex, ages, stages, and tumor grades (data not shown).

Comparing the SHP1-P2 methylation levels for LN N0 and LN– Nx, the percent SHP1-P2 methylation yielded a maximal area under the curve at 0.637. This value could be used detect occult carcinomas with a cutoff value of 10.14%, sensitivity of 66.67%, and specificity of 60% (Figure 1C).