A biomarker is defined as ‘a biological molecule found in blood, other body fluids, or tissues that is a sign of a normal or abnormal process, or of a condition or disease’ by the National Cancer Institute ¹ . The biomarker, also called a molecular marker, has a wide range of applications in diagnosis, monitoring of treatment, and the prognosis of a disease or condition. Despite attempts to classify biomarkers of cancer, a consensus has not been established. Mishra and Verma ¹ have suggested that the classification of biomarkers can be based on the disease state, biomolecules, or other criteria.(Fig. 1)

A peer-review analysis by the World Health Organization International Agency for Research on Cancer (WHO IARC) reported that the global estimated rate for oral cavity cancer was 2.7 per 100,000 in 2012 ^2,3 . Oral cancers are also the eighth most common causes of cancer-related deaths worldwide ⁴ . Oral squamous cell carcinoma (OSCC) accounts for over 90% of oral cancers and is considered to be a rising global public health issue because of its high incidence and low survival rate ^5,6 .

In the efforts to reduce OSCC-related mortality, enhancing and innovating screening and early detection technologies has been suggested as the most effective and fastest developing strategy ⁷ . In this area, the liquid biopsy came up as a noninvasive diagnostic technique that is based on the detection of tumor markers in body fluids. In addition to blood, saliva has a role as an auxiliary tool in oral cancer diagnosis ⁸ . Furthermore, recent studies have revealed that saliva sampling can be a more effective method of detecting specific OSCC biomarkers ⁹ .

Recently, diagnostic markers are the focus of our clinical and experimental studies. A diagnostic cancer marker can be specific to stage, tissue, relapse, follow-up, or age ⁹ , and present at any stage during cancer development. This review article introduces an updated list of reported OSCC salivary biomarkers up until 2019 and discusses the current clinical application of salivary biomarkers.

Whole saliva also contains a variety of non-organic and organic substances from the serum, gingival crevicular fluid, as well as oral microorganisms and their products^10,11. In addition to a diversity of biomarkers for many diseases, saliva’s collection is noninvasive and convenient, and the transportation and storage are easy, therefore saliva sampling is cost-effective and efficient ¹² . These advantages demonstrate that saliva is a potential body fluid for laboratory tests compared to serum and tissue samples.

Biomarkers are detected and determined by various molecular techniques. For the genomic biomarkers (including DNA, mitochondrial DNA [mt.DNA], RNA, messenger RNA [mRNA], microRNA [miRNA]), the utilized techniques can be DNA microarrays, polymerase chain reaction (PCR), Southern blot analysis, restriction fragment length polymorphism (RFLP), and cross-linking immunoprecipitation (CLIP). The proteomic biomarkers class includes proteins, peptides, antibodies, and can be analyzed by liquid chromatography, Western blot analysis, protein sequencing, protein arrays, and immunofluorescence. The metabolomics biomarkers (including carbohydrates, enzymes, metabolites, liquids) are determined by liquid chromatography, nuclear magnetic resonance, enzyme assays, and mass spectrometry ¹¹ . One of the earliest developed saliva biomarkers, human papillomavirus (HPV) markers, have been used as the diagnostic biomarker in the cervical cancer screening program and vaccine development ¹ .

The identification of reliable salivary biomarkers for the OSCC screening has been enhanced thanks to the easy and noninvasive collection of saliva compared to the drawing of blood ⁹ . The underlying tissue changes in the disease process can be classified as genomic, proteomic, or metabolomic expression.(Fig. 2) With the development of salivaomics, a lot of researches have been performed and more than 100 potential saliva biomarkers for OSCC have been reported in the literature ¹³ . Salivary diagnostic has been an attractive potential modality screening, early detection and prognosis evaluation for researchers and clinicians ¹⁴ .

The classification of biomarkers can be based on the disease state, biomolecules, or other criteria ¹ . Currently, we are paying attention to the screening and early detection of OSCC, and diagnostic markers are in the focus of our clinical and experimental studies. Diagnostic markers may be present at any stage of cancer development.(Table 1) The expression of Eph and/or ephrin are common in various primary tumors, including OSCCs. Among the various biological functions of ephrin and Eph receptors in cancer, their involvement in EFNB2/EphB4 signaling is thought to be associated with angiogenesis, differentiation, and development. Therefore, EFNB2 gene expression is suggested to be a useful biological marker in prognostic evaluation in patients with OSCC ¹⁵ . Shpitzer et al. ¹⁶ reported that the levels of 8-oxoguanine DNA glycosylase, phosphorylated-Src, and mammary serine protease inhibitor (Maspin) were found to decrease in the saliva of OSCC patient. Several studies reported that interleukin (IL)-6 and IL-8, which are well-known as post-inflammatory cytokines, significantly increase in patients with OSCC and therefore suggesting their value as a diagnostic marker of oral malignant and premalignant lesions ^17,18 . Arellano-Garcia et al. ¹⁹ reported that both IL-8 and IL-1β were found to significantly increase in OSCC patients.

A diagnostic cancer marker can be specific to tissue, stage, follow-up, relapse, and age. Despite the attempts made to classify cancer biomarkers, a consensus has not been established. However, for the diagnosis and comprehension of the OSCC genomic architecture, more recent efforts have focused on new and noninvasive methods using human saliva sampling, which include proteomic (Table 2)^20-25, proteins (Table 3)^20,25-36, transcriptomic (Table 4)^20,37-41, and metabolomic (Table 5)^20,42-44 biomarkers.

Laboratory examination is an essential and high accurate method for disease diagnosis and prognosis. Among the available laboratory test, liquid biopsy is a less invasive method that limits the need for acquiring tissue ⁴⁵ . In the past 20 years, liquid biopsy has become an essential examination in multiple areas of oncology, based on the detection of circulating tumor cells (CTCs), circulating tumor DNA (ctDNA), and circulating tumor RNA (ctRNA), proteins, and exosomes ⁸ . Liquid biopsy samples include blood, urine, saliva, and other bodily fluids such as seminal plasma, pleural effusions, cerebrospinal fluid, sputum, and stool samples ⁴⁶ .

Saliva has complex components that originate from the major and minor salivary glands, as well as from the oropharynx, gastrointestinal reflux, gingival crevicular fluid, and blood. The analysis of saliva components has been considered an effective method for monitoring the status of health ⁴⁷ because changes in the compounds that constitute saliva can reflect the physiological and pathological status of the body.

1. Blood biomarkers for OSCC

Currently, the most common liquid biopsy is blood. Approximately, 5 to 10 mL of blood is all that is needed for a liquid biopsy. Blood biomarkers are classified as CTCs, ctDNA, ctRNA, proteins, and exosomes, which can be used for differential diagnosis, prognosis, cancer recurrence detection, tumor evolution monitoring, and treatment efficacy in various types of tumors^8,48. Circulating cell-free DNA (cfDNA) is released into the bloodstream from apoptotic or necrotic cells. cfDNAs can originate from nonmalignant host cells and tumor cells, thus including ctDNA. There are also authors suggesting that tumor exosomes can play an important role in immune suppression and enhancing tumor development, and plasma exosomes can be the next generation of biomarkers in head and neck cancer progression evaluation⁸.

2. Standard saliva collections

There are a lot of methods are for the whole saliva collecting, such as draining, spitting, suction, and swabs⁴⁹. A variety of commercial devices and methods for collecting saliva from individual glands have also been developed^20,49.(Table 6) The clinician should be properly trained on performing saliva sample collection to achieve the best performance and samples. Despite the variety of choices, only one type of collection device should be used in one study^50,51. The eligible participants need to be educated and given the appropriate instructions before the collecting procedure. The sample volume needs to be sufficient, and the type of container needs to be prepared beforehand accordingly. Besides, sample labeling and handling protocols must be well-designed and carried out consistently⁵².

Saliva components can vary or remain stable at age⁵³. It is reported that children and adults also can have differences in the salivary level of a specific protein, peptide, and proteome⁵⁴. The unstimulated saliva secretion was higher in healthy men compared with women⁵⁵, indicating gender-dependent secretion. Due to these variables, patients should be categorized into different age groups and gender to minimize statistical errors.

The position of the mouth during saliva collection is important from the basis of salivary glands location and their secretion patterns⁴⁹. While saliva secreted from the major salivary glands contains common components, the concentrations of each component and some specific contents can vary from one gland to another. On the other hand, components of the saliva from the minor glands mainly include mucins and lipase⁵¹.

Besides, saliva components and origin depend on the resting state or stimulated state of the individual. For example, cortisol, alpha-amylase, and secretory IgA levels in the saliva are affected by stimulation⁴⁹. The complex in the composition of saliva and its dependence on various conditions should be considered and evaluated thoroughly by the investigator. This consistency in sampling procedure is essential to achieve valuable data.(Table 7)

3. Saliva preservation with detailed information

The saliva samples can be kept at room temperature for a maximum of 30 to 90 minutes for the immediate analysis⁵⁶. Thomadaki et al.¹⁰ recommended that the samples are frozen at or below –20ºC immediately after collection, due to the low temperature can slow down the degradation of the salivary proteome. Salivary specimens can also be stored at –80ºC for several years with little or no degradation⁵⁶. In RNA analysis sampling, an RNase inhibitor should be added in the supernatant fractions before storing it at –80ºC⁵⁷. Shirtcliff et al.⁵⁸ suggested that the collection method was an important cause of the unsystematic error. Samples obtained by spitting contain more bacteria than drooling samples, which can affect the analysis results of saliva compounds⁵⁶.

As with saliva, blood is a complex fluid that contains a wide range of molecular components, including various antibodies, growth factors, enzymes, and hormones. Therefore, blood serum and plasma are the traditional and conventional sources of liquid biopsy and biomarker examination. Although used widely, blood sampling and analysis can often be expensive, problematic, and invasive.

Comparatively, sampling saliva has many advantages over blood, including the following ⁵² : 1) The saliva sampling procedure is easier and can be self-collected, 2) the procedure is noninvasive, 3) samples are safer to handle and 4) are easier to transport and store because saliva requires simpler manipulation than blood, and 5) the procedure is cost-efficient both for patients and the investigators.

Despite these advantages, the use of saliva sampling as a diagnostic tool is still under controversy. The greatest obstacle in the development of a salivary diagnostic protocol is that while most biomarkers detected in the blood serum are also can be found in saliva, but their levels are so low that they barely can contribute to the diagnosis. For example, the normal level of IgG (5 to 30 mg/mL vs 5 to 30 μg/mL) and IgM (0.5 to 1 mg/mL vs 5 to 10 μg/mL) also have a huge difference in serum and saliva analysis ⁵⁹ . However, the link between blood and saliva at a molecular level may also suggest that saliva can be a potential alternative to blood- and tissue-based diagnostics.

While there are reports about many saliva biomarkers, the correlation of their levels in the blood and saliva are various, and there are a scarce number of publications on the blood levels of specific biomarkers found in saliva, or how the saliva collection technique affects the quality and diagnosis value of specific biomarkers. Williamson et al. ⁵⁹ analyzed the correlation of biomarkers among passive drool saliva, filter paper saliva, and plasma samples in healthy adults. The author found that between passive drool and filter paper saliva samples, statistically significant correlations were found among 16 cytokines, including IL1β, IL-1ra, IL-4, IL-7, IL-8, IL-9, IL-10, IL-12, IL-13, IL-15, granulocyte colony-stimulating factor (G-CSF), interferon gamma (IFN-γ), IFN-γ-inducible protein 10 (IP-10), monocyte chemoattractant protein 1 (MCP-1), macrophage inflammatory protein-1β (MIP-1β), and vascular endothelial growth factor (VEGF). When plasma and passive drool saliva samples were compared, only 3 cytokines, IL-6, IFN-γ, and MIP-1β, were statistically significantly correlated ⁴⁹ . Glutathione (reduced glutathione [GSH] and oxidized glutathione [GSSG]) is an anti-oxidant that responds to both xenobiotic and endogenous compounds. Ngamchuea et al. ⁶⁰ found a weak correlation between salivary and whole blood glutathione content (GSH+2GSSG) in healthy participants. Almadori et al. ²⁶ suggested that salivary glutathione levels may be an index of oxidative stress at the level of the upper airways and in particular of the oral cavity and pharynx. A study by Sharma et al. ⁶¹ showed a significant and gradual increase in serum and salivary L-fucose between control and oral potentially malignant disorders (OPMDs) or oral cancer (OC) samples. The authors suggested that L-fucose can be used as a reliable biomarker and saliva can be used as a diagnostic fluid for the screening and early detection of OC. A significant positive correlation was found between serum and salivary Cyfra 21-1, serum Cyfra 21-1, and CK19 mRNA expression and between salivary Cyfra 21-1 and CK19 mRNA expression ⁶² . The magnesium concentration was low in both the blood plasma and saliva of OSCC when individuals with potentially malignant disorders were compared to healthy subjects. Thus, the magnesium ion concentration in blood plasma and saliva could be considered as a tumor marker, playing an important role in carcinogenesis ⁶³ . Oral pre-malignancy and malignancy patients were reported to have lower serum albumin levels compared to healthy individuals. Otherwise, salivary albumin levels were found to increase in oral pre-malignancy and oral malignancy cases compared to healthy individuals. A study by Metgud and Patel ⁶⁴ suggested that albumin may play a role in early diagnosis and prognosis of oral premalignant and malignant tissues.

Salivary biomarkers are a very promising noninvasive approach to oral cancer detection. In monitoring the disease progression and patient’s response to therapeutics, salivary biomarkers show many advantages because of the noninvasive and cost-effective sampling methods. However, the current challenges in salivary biomarker research are standardizing sample collection, improving sample processing and storage, and reduce the wide variability in cancerous and non-cancerous individuals. Useful biomarkers in the screening and early diagnosis of OSCC are still under research but will be defined in the near future.