Clinical implications of heat shock protein 70 in oral carcinogenesis and prediction of progression and recurrence in oral squamous cell carcinoma patients: a retrospective clinicopathological study

Oral cancer makes up to approximately 4–5% of all cancers globally. Oral cancer has a high incidence rate due to various risk factors, including tobacco and alcohol use, immune defects, genetic factors, and viruses [1]. Oral cancer is considered a common cause of death with an incidence rate of about 400,000 cases, and 223,000 deaths annually worldwide [2, 3]. Death is usually observed in advanced stage of the disease [4]. About 90% of oral neoplasms and 38% of head and neck tumors are oral squamous cell carcinomas (OSCC) [5]. Approximately 4500 persons are diagnosed with oral cancer each year in Egypt, where OSCC accounts for almost 90% of all oral malignancies [6].

Regardless of the great advances in diagnosis and management, the 5-year survival rate of OSCC patients is about 50% even following successful treatment intervention. long-term prognosis is compromised by the late initial presentation of advanced tumors, tumor metastasis, and subsequent recurrence [7, 8]. Several factors have been reported for predicting the prognosis of OSCC, including clinicopathological features, e.g., clinical stage, tumor differentiation, tumor size, and treatment types, etc. [9‐11]. Tumor size, cervical lymph node metastases, and recurrence are the main clinical prognostic factors [12‐14].

One of the most prevalent and evolutionary highly conserved intracellular compounds is heat shock protein (HSP) [15‐18]. Both prokaryotes and eukaryotes contain several copies of the HSP70 protein. Thirteen HSP70 homologs have been identified in human cells, including the cytosol, nucleus, lysosomes, extracellular space, and mitochondria, pointing to unique and organelle-specific biological functions [18]. Under physiological settings, this class of proteins maintains the vital cellular processes and shields cells against a variety of challenges, including ischemia, infections, inflammation, heavy metals, and hyperthermia [19]. Long lines of experimental data point to HSP70’s critical function in cancer [20]. It frequently acts as a biomarker for bad prognosis and is highly expressed in malignant tumors [21, 22]. HSPs were associated with the risk, progression, and prognosis of oral cancer. However, no consistent conclusion could be acquired from the literature because of the contradictory results. Additionally, there were great conflicts in the findings concerning the expression of HSP70 in cancers. Many studies concluded a significant association between increased expression and poor prognosis in cancer patients, while others presented contradictory results. For that reasons, the current study was conducted to elucidate the role of HSP70 in oral carcinogenesis by investigating its expression in different tissue levels ranging from normal oral mucosa to dysplastic oral epithelium and OSCC. Moreover, HSP70 immuno-expression was evaluated concerning different prognostic parameters to determine its capability in predicting cancer progression. Recurrence of the tumor was recorded, and patients` disease-free survival was calculated and analyzed considering HSP70 expression to determine the potential utility of HSP70 immuno-expression in predicting recurrence.

The current retrospective study was carried out on 50 cases of OSCC. Biopsies from the cancerous tissue along with the free surgical margin on the periphery of the tumors were included in the study. Five specimens of the normal oral mucosa were used as a control group.

In numerous biological processes, such as cell differentiation, gene expression, immune system control, cellular senescence, and programmed cell death, HSPs play a crucial role [27‐32]. HSPs are essential for invasiveness, cell proliferation, angiogenesis, and cancer development [33, 34]. They are involved in the development and spread of tumors [35]. Numerous physiologic and pathological processes, including immunoregulation, apoptosis, tumor treatment resistance, oncogene activation, and tumor suppressor gene inactivation have been linked to HSPs [36‐41]. Numerous tumors, including colorectal, breast, prostate, lung, ovarian, gastric, oral, and esophageal malignancies, have been found to express HSPs abnormally [23, 42]. In the current study, HSP70 expression has been evaluated in a spectrum of normal, dysplastic oral epithelium, and OSCC to elucidate its contribution to oral carcinogenesis. Additionally, the expression was analyzed concerning the tumor histologic grade of differentiation, and patients` parameters like sex, tumor site, and age. Moreover, HSP70 expression was evaluated in relation to different tumor sizes, the incidence of nodal and distant metastasis, and finally the incidence of tumor recurrence.

The conducted study presented differential expression of HSP70 in different tissues (normal oral mucosa, dysplastic oral epithelium, and OSCC). Negative HSP70 expression was noted in all the worked cases of normal oral mucosa, while the dysplastic oral epithelium presented mainly low expression in contrast with the carcinoma group that presented mainly high HSP70 expression. In the same line as our work, many studies reported the same finding; Kaur et al. [43] reported increased levels of HSP70 expression in oral cancer than in potentially malignant lesions. The authors deduced the implication of HSP70 in oral carcinogenesis depending on that finding. Moreover, Markopoulos AK et al. [44] assessed the expression of HSP70 in OSCC, Leukoplakia with dysplasia, and normal oral tissue. They also documented Intensified expression of HSP70 in OSCC compared with dysplastic lesions that presented only positive expression, and normal oral mucosa that showed negative expression. They advised the utility of HSP70 as a marker for epithelial dysplasia or epithelial malignant transformation. Oral cancer risk was substantially correlated with HSP70 overexpression [45]. Furthermore, Tekkesin et al. [46] recorded increased expression of HSP70 and HSP27 during tumorigenesis in OSCC, and Thubashini et al. [47] also observed Increased HSP70 expression from oral submucous fibrosis to OSCC. In a study carried out on nasopharyngeal carcinoma (NPC), cancerous tissue showed considerably higher levels of HSP70 expression in comparison to the non-cancerous control nasopharyngeal epithelium [48]. Lung cancer studies also registered a relatively higher risk of lung cancer with increased expression of Hsp70 [49, 50]. Induction of HSP70 expression usually occurs in response to stress [43]. This fact explains the higher level of expression in dysplastic oral epithelium and oral cancer than in normal oral mucosa.

In contradiction, Sugerman et al. [51] found that HSP70 is not a definitive marker of oral malignancy in a study carried out on OSCC, dysplastic oral epithelium, and normal oral epithelium. The contradictory findings may be owing to miss understanding of the molecular mechanisms responsible for the overexpression of heat shock proteins in cancer cells as it may be tumor specific [51]. The physiopathological characteristics of the tumor microenvironment (low glucose, pH, and oxygen) are one of the processes that cause HSP induction [52]. The other mechanism is the oncoproteins that appear during carcinogenesis as mutated P53. Kaur et al. [43] proposed that p53-HSP70 complex formation may be one of the mechanisms of stabilization of p53 protein resulting in its increased levels in potentially malignant and malignant oral lesions and may be implicated in oral carcinogenesis.

HSP70 immuno-expression not only presented differential expression along the different tissue levels but also there was significant differential expression along the three histologic grades of dysplastic oral epithelium in the free surgical margin group, where HSP70 expression was significantly intensified by progression from mild to moderate and severe dysplasia. HSP70 expression is exaggerated by acquiring the malignant characteristics of cells. Moreover, high HSP70 expression was significantly observed in moderately and poorly differentiated carcinoma groups than in the well-differentiated carcinomas group that presented (mainly) weak expression. Similar conclusions were drawn from other research, including the high levels of HSP70 expression seen in moderately and poorly differentiated carcinomas [53] and elevated levels of expression linked to carcinogenesis [54, 55]. HSP70 expression and the differentiation of oral cancer yielded conflicting results as well [56, 57]. HSP70 expression did not significantly differ across the distinct histologic types of nasopharyngeal carcinoma (differentiating and undifferentiating types) [48]. It was important to note that oral cancer development can be accompanied by varying levels of HSP70 protein production, which might account for the inconsistent results.

Regarding the clinical variables; females in our study significantly presented a predominant high HSP70 expression than males (94.7% versus 71%, respectively). The gender of the patient had a statistically significant impact on the expression of HSP70, according to Pearson’s Chi-square test (P = 0.041). The female sex hormones may be responsible for that difference. Opposing our findings, Thubashini et al. [47] observed no association of HSP70 expression with a patient`s gender in an oral cancer study. Moreover, Feng et al. [48] reported similar findings concerning patient gender and age also in nasopharyngeal carcinoma. Furthermore, high HSP70 expression was recorded in the Japanese population among male patients over females in lung cancer studies [50, 58]. The opposing findings may be owing to factors related to special characteristics of the worked samples, or different environmental factors.

The worked study revealed no statistically significant differences in HSP70 expression neither concerning the patient`s age group (P = 0.544) nor the varied intraoral sites of the tumor (P = 0.280). Many studies reported findings on the same line as ours [48, 49, 54]. In contrast, Thubashini et al. [47] observed Increased HSP70 expression in ages older than 50 years in OSCC patients.

HSP70 expression was strongly correlated with OSCC progression and was seen in all cases with large-sized tumors (T3 and T4, P = 0.002, advanced TNM clinical stages (III and IV, P = 0.001), positive nodal involvement, and high expression (P = 0.002). HSP70 expression was high in all M1 stage cases, although there was no statistically significant relationship between HSP70 expression and the incidence of distant metastasis (P = 0.077). On the same line as our findings, HSP70’s high level of expression was significantly associated with advanced clinical stages (III and IV), the presence of nodal metastasis, but contradicts our finding regarding metastasis as high expression was associated with the presence of metastasis in nasopharyngeal carcinoma [48]. In contrast, Lee et al. [56] found that low HSP70 expression was associated with lymph node metastasis in OSCC patients.

Consistent with our findings concerning the size and the clinical stage of the tumor, non-small cell lung cancer patients with histology of adeno and squamous cell carcinoma with significant gross tumor volume had high serum Hsp70 as compared to normal healthy controls. Additionally, there was a difference in HSP70 that was noted in serum samples between patients with early-stage and advanced-stage tumors [59]. High HSP70 expression associates with advanced clinical stage in OSCC studies [53, 54]. Patients with HSP70 overexpression exhibited a more advanced stage of oral cancer, according to Zheng et al. [60]. However, according to Taghavi’s study, HSP70 overexpression was discovered to be a protective factor for the advanced stage [57].

Kaur et al. [53] worked on human oral squamous carcinoma cells (HSC-2 cell line) and deduced the implication of HSP70 in the proliferation and survival of oral tumor cells by the anti-apoptotic function of HSP70. This finding explains its increased expression in cancer cells than normal controls, in moderately and poorly differentiated carcinomas that are characterized by high proliferation capability than well-differentiated carcinomas, and in large-sized tumors than small-sized tumors. HSP70 participates in increasing the size of tumors via suppressing apoptosis pathways. Continuously under stress, such as hypoxia and nutritional depletion, tumor cells can survive but are vulnerable to apoptosis [61]. Cancers frequently contain the protein HSP70, which has been shown to block apoptosis and promote cancer growth [62]. Notably, a reduction in HSP70 level was found to cause the death of cancer cells [63]. Both intrinsic and extrinsic apoptotic pathways were shown to be affected by HSP70. By inhibiting c-Jun N-terminal kinase (JNK) and p38, HSP70 prevents apoptosis through a caspase-dependent mechanism. JNK activity is crucial for the release of cytochrome c from the mitochondria, which triggers the intrinsic apoptotic pathway [64]. The other mechanism is caspase-independent through the interaction with death receptors and the suppression of the apoptosis-inducing factor (AIF) [65]. The apoptosis-inducing factor (AIF) necessary for DNA fragmentation is inhibited by HSP70 [66]. HSP70 is a key player in necrosis and autophagy, two alternate forms of cell death to apoptosis. HSP70 was discovered to be present in the lysosomes of cancer cells, in contrast to normal cells [67]. Numerous investigations have shown that HSP70 stabilizes lysosomal membranes, preventing cancer cells from death [67‐69]. Apoptozole, a HSP70 inhibitor, has demonstrated its efficacy in encouraging apoptosis in cancer cells by inducing the release of cathepsin and permeabilization of the lysosomal membrane [70].

Similar to our findings, high expression of HSP70 has been linked to lymph node metastases in breast cancer models [71]. Inactivation of HSP70 was reported to diminish tumor invasiveness and metastatic potential in breast, cervical, and bladder cancer cell lines in subsequent studies [72]. High levels of HSP70 expression shield tumor cells from anoikis and amorphosis, two cell death processes that take place when the extracellular matrix (ECM) is lost [73, 74]. HSP70 was found to have an impact on molecules related to anoikis and amorphosis, including Akt and focal adhesion kinase (FAK) [75‐77]. HSP70 stabilizes the buildup of hypoxia-inducible factor-1 (HIF-1), which is implicated in tumor angiogenesis, invasion, and metastasis [78‐80] and senses low oxygen levels. Extracellular HSP70 has been demonstrated by Kim and colleagues to attach to the surface of endothelial cells and trigger angiogenesis through an ERK-dependent mechanism [72]. Additionally, it has been noted that HSP70 increases IL-5-induced angiogenesis, pointing to an inflammatory role for HSP70 in encouraging angiogenesis [70]. Furthermore, numerous studies have shown that HSP70 blocks TGF-induced EMT by reducing Smad2 phosphorylation [81‐83]. Furthermore, it has been demonstrated that HSP70 inhibits the formation of reactive oxygen species (ROS) and Smad3 and Smad4 phosphorylation, thereby protecting peritoneal mesothelial cells against high glucose-induced EMT [84]. These results point to a second function for HSP70 in EMT, where HSP70-peptide complexes aid tumor cells’ shift to mesenchymal phenotype. MMP-9, a processing enzyme for cleaving the CD44 receptor involved in adhesion and migration, is released as a result of extracellular HSP70 [85‐87]. Additionally, Wiskott–Aldrich syndrome family member 3 (WASF3), a protein necessary for tumor migration, invasion, and metastasis, was shown to be stabilized by HSP70 and HSP90 proteins [87‐89].

Utilizing the Kaplan–Meier method, log-rank test, and Cox regression model to analyze the Patients’ DFS in relation to various clinicopathologic factors. In moderately and poorly differentiated carcinomas, carcinomas with high HSP70 expression, large-sized tumors, positive nodal involvement, positive distant metastases, and advanced TNM stage, the DFS of patients was significantly reduced, according to a univariate study. However, there was no statistically significant difference between DFS and the patient’s age (P = 0.338), gender (P = 0.310), or intraoral tumor site (P = 0.053). Nodal involvement and tumor size were discovered to be the independent predictors of patient DFS in a multivariate study utilizing the Cox regression model. In harmony with our finding, Kaur et al. [53] reported elevated levels of HSP70 associated decreased median disease-free survival time. In other work on human oral squamous carcinoma cells (HSC-2 cell line) they concluded that HSP70 is required for the proliferation and survival of oral tumor cells [52]. Moreover, many studies reported HSP expression was inversely correlated with survival time [54, 90]. Based on the findings of non-small cell lung cancer studies, HSP70 overexpression can be used to predict survival [91].

Contrarily, numerous studies have linked lower HSP expression with a poor prognosis for oral cancer patients [52, 55, 90]. According to Lu et al. HSP70 protein overexpression significantly increased the survival rate of patients with oral cancer [45]. A small number of other investigations found no predictive association between HSP expression and patient outcome [92]. When Ito T et al. [46] examined the tongue, normal epithelium, dysplasia, and cancer, they discovered no connection between the expression of HSP27, 60, 70, 90, or p53, and survival. Positive clinical outcomes for gastric and colon cancers supported the hypothesis that natural killer cells (NK) should target mHSP70, while lower rectal and squamous cell carcinomas had a bad prognosis, according to HSP70-membrane positive phenotype (mHSP70) research [93]. The authors concluded that the variations in these malignancies' metastatic paths can account for the discrepancies in clinical outcomes [93]. HSP70 overload causes immune tolerance to tumor antigens, which aids in the progression of tumors, while the initial release of HSP70 acts as a tumor suppressant that causes the killing of membrane-positive HSP70 tumor cells by NK cells to explain conflicting findings regarding the expression of HSP in cancer.

Heat shock proteins may not currently be on the list of helpful prognostic indicators in oral cancer due to these inconsistent results.

The significant differential expression of HSP70 at different tissue levels (normal, dysplastic epithelium, and OSCC) elucidates its contribution to oral carcinogenesis. Moreover, high HSP70 expression is significantly associated with poor prognostic factors. The level of HSP70 expression in cancerous tissue is predictive for the severity and progression of cancer. Therefore, HSP70 immune-expression could potentially be used as predictor of progression and recurrence of OSCC patients.

Although immunohistochemistry offered valuable findings in measuring HSP70 expression at different tissue levels, using other techniques such as western blotting, in situ hybridization, or PCR could enrich the work. Furthermore, while the worked sample is regarded limited to some extent, it is still valid in providing useful insights. As a result, I recommend working on larger sample in the future studies to could generalize the conclusions.