LZTS2 promoter hypermethylation: a potential biomarker for the diagnosis and prognosis of laryngeal squamous cell carcinoma

Laryngeal cancer is one of the most common head and neck tumors and has the second highest mortality rates of all respiratory system malignancies [1]. More than 95% of laryngeal carcinomas exhibit laryngeal squamous cell carcinoma (LSCC) pathological features [2]. The pathogenesis of LSCC is complicated and involves genetic, epigenetic, and environmental factors, such as alcohol, tobacco, and asbestos [3]. While recent advancements have been achieved in therapeutic strategies that combine surgery, chemotherapy, and radiotherapy, the 5-year overall survival rate of LSCC remains poor [1, 2].

DNA methylation is a well-studied mechanism driving epigenetic regulation of gene expression [4]. In addition to mutations and deletions, the abnormal methylation of tumor suppressor gene (TSG) promoter is considered to be the third mechanism of inactivation of TSG [4], which mainly happens in the CpG island near the original site of the gene transcription and can lead to the abnormal expression of TSG. Numerous studies have shown that hypermethylation of CpG islands in TSG plays a crucial role in carcinogenesis and progression in various solid and liquid tumors and can be a molecular marker used to identify LSCC [4], such as acute myeloid leukemia [5], bile duct carcinoma [6], and breast cancer [7].

LZTS gene family members share functions related to transcription regulation and controlling the cell cycle [8]. LZTS2 is located at human 10q24.3 (Fig. 1), which is proximate to 10q23.3, the site of the prototypical tumor suppressor gene PTEN [9]. Previous studies have shown that expressions of both regions are frequently downregulated in a variety of tumors, suggesting that other tumor susceptibility genes may exist in these areas in addition to PTEN [10]. Mounting evidence supports that LZTS2 regulates cell growth through protein-protein interactions with β-catenin [11]. β-catenin, one of the subunits of the cadherin protein complex, also performs critical functions in the Wnt signaling pathway. Accumulation of β-catenin in the nucleus plays a vital role in tumorigenesis and progression [12]. LZTS2 interacts with β-catenin to both repress its transcriptional activity and regulate its subcellular localization and signaling; together, these activities suppress cell growth [11]. In addition, Johnson et al. demonstrated, in LZTS2 knock-out mouse embryonic fibroblasts, that a lack of LZTS2 expression promoted cell survival and proliferation [10]. Thus, evidence supports that LZTS2 is a tumor suppressor gene and that aberrant expression contributes to the genesis and development of some tumors [10].

To date, no reports have investigated methylation of the LZTS2 promoter in human cancers, including LSCC. Therefore, the aim of the present study was to explore whether correlations exist between LZTS2 promoter methylation and risk of LSCC.

LZTS2 promoter methylation levels were evaluated in 96 LSCC specimens and paired adjacent normal tissues using qMSP technologies. The present data revealed that LZTS2 promoter methylation levels were significantly higher in LSCC cancer tissues when compared to paired adjacent normal tissues (Fig. 2, P = 1.37e⁻⁰⁶). Given that smokers are at risk of developing LSCC [14], the cohort was stratified according to smoking status. Subgroup analyses revealed that the observed LZTS2 hypermethylation patterns occurring in LSCC (when compared to paired normal tissues) were more clearly evident in smokers than non-smokers (Fig. 2, smoking P = 4.74e⁻¹⁸; non-smokers P = 0.029). Next, associations between clinicopathological characteristics (e.g., age, gender, smoking behavior, histological classification, T classification, clinical stage, and lymph node metastases) and LZTS2 promoter methylation rates were further explored (Table 1). As expected, the subgroup analyses confirmed that LZTS2 promoter methylation levels of LSCC in smokers were significantly elevated, relative to non-smokers (P = 0.049). Additionally, T classification, clinic stage, and presence of lymph metastases all served as significant explanatory variables. The PMR in advanced clinical stages (stages III + IV) and T classifications (T III + IV) were significantly greater than specimens exhibiting earlier clinical stages (stages I + II; P = 0.005) and lower T classifications (T I + II; P = 0.003). In contrast to patients without lymph node metastases, methylation levels of patients with lymph node metastases were significantly elevated (P = 0.019). The mean of PMR in male subjects, subjects over 60 years old, and subjects with well-differentiated histologies were numerically higher than female subjects, subjects under 60 years old, and subjects with moderately and poorly differentiated histologies, respectively; however, these differences were not statistically significant (Table 1).

To evaluate the potential diagnostic value of LZTS2 methylation, the receiver operating characteristic (ROC) curve was plotted. This technique served to identify a diagnostic threshold value for LZTS2 promoter methylation. The area under the ROC curve (AUC) was 0.920 (Fig. 3). The diagnostic threshold value (cut-off value) for LZTS2 methylation was 11.63% (94.7% sensitivity and 80.4% specificity). The values over the cut-off were defined as positive diagnostic indicators, while those below the cut-off were considered negative indicators. The false positive and false negative rates were 18.8 and 5.2%, respectively. The positive predictive value was 83.5% and negative predictive value was 94.0%. The diagnostic accordance rate was 88.0%.

To evaluate whether an association existed between patient overall survival and LZTS2 methylation, survival analyses were performed on two subcohorts that were split at the average LSCC PMR value (average value: 0.2706). Fifty-nine patients were classified into the hypomethylation group and 37 were assigned to the hypermethylation group. Using Kaplan-Meier analyses, overall survival of LSCC patients with hypomethylated LZTS2 promoters was significantly higher than that of patients with hypermethylated LZTS2 promoters (Fig. 4, log-rank test P = 0.028). Univariate Cox proportional hazards analysis also revealed an obviously increased risk of death for LSCC patients with hypermethylated LZTS2 (HR = 44.366; 95% CI = 4.586–429.237; P = 0.001). Subsequently, we performed a multivariate Cox proportional hazard analysis by adjusting for smoking behavior, histological differentiation, clinical stage, and lymphatic metastasis. The results confirmed that LZTS2 promoter hypermethylation could be an independent factor to predict a poorer overall survival of LSCC patients (HR = 6.671; 95% CI = 2.087–21.324; P = 0.001, Table 2).

For LSCC patients, early-stage disease often goes undetected because symptoms are mild and non-specific. Consequently, roughly 40% of newly diagnosed patients present with stage III or IV disease [15]. Currently, the prevailing initial therapy for LSCC is total laryngectomy combined with postoperative radiotherapy. Given that the larynx plays an indispensable role in human communication and that the operative treatment damages laryngeal structure and function, current treatment paradigms greatly reduce patients’ quality of life [13]. Patients for whom disease is identified in T1 or T2 stages have an 80–90% cure rate; for advanced LSCC, cure rates drop to about 60% [2]. Therefore, novel methods are needed to allow for earlier identification and possible prevention of LSCC, while cancers remain curable and laryngeal function can be best preserved.

Abnormal TSG promoter methylation is an early and frequent event occurring in LSCC tumorigenesis [4]. Accumulating evidence indicates that hypermethylated gene promoters exist in LSCC tissue, as well as in the serum and saliva of LSCC patients [4]. It stands to reason that hypermethylation of a panel of TSGs could serve LSCC patients as sensitive diagnostic and prognostic biomarkers, or as markers following the natural history of curative therapies [4].

Here, a novel TSG was evaluated. LZTS2 is located at human 10q24.3, is expressed in most normal tissues [8], suppresses cell growth by interacting with β-catenin [11], and plays a vital role in the carcinogenesis and development of tumors [10]. To date, research has focused on the LZTS2 TSG functions; however, the relationship between LZTS2 promoter methylation events and tumorigenesis was largely unstudied. This study investigated associations between LZTS2 promoter methylation and risk of LSCC, as well as LZTS2 clinical diagnostic value. Previous studies have shown that LZTS2 cDNA overexpression suppressed cell growth and reduced colony formation rates of several cancer cell lines, including HEK-293T, AT6.2, LNCaP, PC3, TRSUPr1, Rat-1, and U2OS [8]. Additionally, LZTS2 protein has low expression levels in human prostate cancer cells [10]. The present results revealed that LZTS2 promoter methylation levels in LSCC tissues were significantly higher than in non-cancerous paired control specimens. Thus, the data suggested that LZTS2 promoter hypermethylation decreased LZTS2 protein expression, which may have regulated LSCC carcinogenic mechanisms and increased risk of developing the disease.

Next, clinicopathologic parameters were identified that served as explanatory variables contributing to LZTS2 methylation levels. Previous studies demonstrated that, in addition to lung cancer, cigarette smoking is a risk factor for upper aero-digestive tract cancers including laryngocarcinomas, hypopharyngeal carcinomas, and esophageal cancers [14]. Additionally, smoking gives rise to extensive genome-wide DNA changes especially DNA methylation, which play important roles in tumorigenesis [16]. Therefore, the current study specifically investigated a relationship between LZTS2 promoter methylation and cigarette smoking. LZTS2 promoter methylation levels in LSCC specimens from smokers were significantly elevated, relative to non-smokers (P = 0.049). Thus, smoking behavior may have elevated LZTS2 promoter methylation levels and thus increased risk of LSCC in those patients. Additionally, LZTS2 promoter methylation levels were significantly increased in patients with lymph node metastases, advanced clinical stages, and advanced T classifications. It is well-understood that T classification and lymph node metastases are crucial to the prognosis of LSCC patients [17]. The present results suggested that LZTS2 promoter methylation also contributed to LSCC progression and ultimate prognosis. Furthermore, previous studies have reported that abnormal TSG methylation was associated with poor survival outcomes in several tumor types [5‐7]. Survival analyses presented here indicated that patients with LZTS2 promoter hypermethylation have worsened outcomes compared to those with hypomethylation in the same promoter regions (Fig. 3), which was consistent with the results of our univariate Cox proportional hazards analysis. Furthermore, the multivariate Cox proportional hazard analysis we performed confirmed that LZTS2 methylation was an independent adverse factor for LSCC outcomes. These findings suggested that LZTS2 promoter hypermethylation could be a potential biomarker for prognosis of LSCC.

The advent of tumor-specific biomarkers has made major strides in advancing clinic diagnostic practices in a variety of tumors, including CEA and CA19-9 for colorectal cancer [18], AFP for hepatocellular carcinoma [19], and PSA for prostate cancer [20]. However, no biomarker currently exists to assist in clinical prognostication of laryngeal cancer. Here, the AUC of the curve of ROC was 0.920; a large AUC is indicative of a high diagnostic value. The sensitivity, specificity, and diagnostic accordance rates were 94.7, 80.4, and 88.0%, respectively. These data provided strong evidence that LZTS2 promoter hypermethylation could be clinically applicable for use in early identification of LSCC.

LZTS2 promoter hypermethylation was associated with LSCC risk, progression, and prognosis, with the strongest associations observed in the subgroup of patients who smoked. Taken together, LZTS2 promoter hypermethylation has the potential to become an LSCC diagnostic and prognostic biomarker.