Although clinical treatments are developing, cancer is still a significant challenge due to recurrence and metastasis. The unbridled proliferation and uncontrollable dormancy have been regarded as two main hallmarks of cancer (Yeh and Ramaswamy, 2015). Recent studies have discovered a direct association between cancer progression and tumor dormancy (Boire et al., 2019). In other words, tumor dormancy is one of the leading causes of tumor outgrowth, relapse, and metastasis. In 1934, tumor dormancy was firstly defined by Willis as an intricate phenomenon that tumor cells stop proliferating but remain capable of malignant progression (Willis, 1953). In recent years, tumor dormancy has been extensively investigated. Tumor dormancy can occur in various phases of tumor progression, which is a primary cause of tumor recurrence even the primary tumor has been resected for decades (Kleffel and Schatton, 2013). Therefore, a deep understanding of the mechanisms of tumor dormancy is needed to develop more promising therapeutic strategies to eliminate dormant cells.

Based on the physiological traits, tumor dormancy can be divided into two categories: cellular dormancy and tumor mass dormancy. Cellular dormancy, a quiescence of solitary cells, is characterized by each cancer cell residing in the G0-G1 phase of the cell cycle. Besides, the underlying mechanisms of cellular dormancy, governed by the extracellular matrix, nutriture, and metastatic microenvironment, are complicated (Dittmer, 2017; Endo and Inoue, 2019). Tumor mass dormancy is characterized by entire neoplasm cells entering a balance between cell division and apoptosis, mediated by angiogenic dormancy and immunologic dormancy (Gao et al., 2017; Hen and Barkan, 2020). With the changes in the surrounding environment, these cessative cells can re-enter the cell cycle, leading to cancer recurrence. Likewise, the progression of cancer occurs in tumor mass dormancy when angiogenic dormancy and immune dormancy are reawakened by surrounding stimuli, after which the balance between proliferation and apoptosis will be broken (Yeh and Ramaswamy, 2015), leading to cancer recurrence. Besides, tumor dormancy is not a single state but an equilibrium determined by cell-intrinsic characteristics and surrounding niche (Senft and Jeremias, 2019).

Stem cells are units of biological organization with clonogenic potential and self-renewal capacity and can be differentiated into multiple functional cell lineages (Dulak et al., 2015). Up to now, stem cells have not been clearly defined due to elusive identities. Therefore, the classifications of stem cells are varied. There are two main classification criteria. In terms of development potential, stem cells can be divided into totipotent stem cells, pluripotent stem cells, and pluripotent stem cells. Totipotent stem cells are capable of differentiating into various kinds of cells and developing into complete human beings (Zakrzewski et al., 2019). Pluripotent stem cells are accompanied with the potential of differentiating into multi-tissue but don’t possess the ability to develop into a complete individual (Abu-Dawud et al., 2018). Unipotent stem cells, derived from further differentiation of pluripotent stem cells, can only differentiate into one or two closely related types of cells (Dulak et al., 2015).

On the other hand, it has been proposed that the main sources of stem cells include embryonic tissues, adult tissues, fetal tissues, and differentiated cells with genetically reprogrammed (Qin et al., 2016). Therefore, based on their origins, stem cells can be divided into four categories: embryonic stem cells, fetal stem cells, adult stem cells, and cord blood stem cells (Li et al., 2020). Embryonic stem cells possess the capability of totipotent stem cells, which have a good potential for differentiation (Dou et al., 2020); Fetal stem cells have a higher potency of multi-differentiation than adult stem cells and can differentiate into numerous cell phenotypes (Bacakova et al., 2018). Adult stem cells are multipotent and are capable of differentiation into only a limited number of cell types, and therefore can only be used to repair and regenerate elected tissue (Han et al., 2019). Cord blood stem cells include hematopoietic stem cells (HSCs) and MSCs (Mohr and Zwacka, 2018). HSCs are generally isolated from bone marrow and have been successfully used in clinical treatment for 50 years, especially in hematological malignancies (Bujko et al., 2019). In recent years, MSCs also have shown great application potential and are mainly applied to treat tumors and various systemic diseases (Pajarinen et al., 2019). Details about MSCs are summarized in Chapter 4.

The research interest in stem cells is due to the potential application in a wide range of clinical treatments, including tissue engineering, cell replacement therapy, gene therapy, system reconstruction, and anti-aging treatment. However, the underlying mechanisms have not been thoroughly studied, which also promotes more researchers to focus on these unknown fields. In recent years, an increasing number of studies have been performed to investigate the interaction between stem cells and tumor development. Especially, with the discovery of cancer stem cells, the roots of uncurable cancer are gradually being ascertained. Among all stem cells, MSCs have been one of the most promising ones in cancer treatment. This review summarizes the mechanisms of MSC-mediated modulation of tumor dormancy and discusses the relationship between cancer stem cells and dormant tumor cells.

In 1997, Bonnet et al. first isolated CSCs from patients of acute myeloid leukemia (Bonnet and Dick, 1997), and discovered that these CSCs show similar characteristics with normal stem cells, which are heterogeneous and also possess the capacity of self-renewal and multi-differentiation (Desai et al., 2019). During the next decades, it has been further confirmed that CSCs are a small subset of cancer cells that are resistant to conventional therapies and can initiate tumor formation (Eltoukhy et al., 2018b). Nowadays, it has been reported that the mechanisms of CSCs formation can be explained in two ways, namely the gene mutations of normal stem cells and the dedifferentiation of progenitor cells (Scheel and Weinberg, 2012; Crea et al., 2015; Eltoukhy et al., 2018b), also showing that CSCs are not a single subpopulation but a group of multi-source cells.

Based on the differentiation capacity, CSCs can be divided into dormancy-competent CSCs (DCCs) and dormancy-incompetent CSCs (DICs) (Talukdar et al., 2019). In this review, DCCs are mainly discussed as being capable of escaping immune clearance and entering dormancy and being responsible for tumor progression, metastasis, chemoresistance, as well as recurrence (Jones et al., 2012; De Angelis et al., 2019). Further studies have discovered that there are many similar characteristics and signaling pathways between DCCs and dormant tumor cells (Talukdar et al., 2019). For instance, it has been reported that DCCs and dormant cancer populations both can maintain quiescence, delay cancer recurrence, resist tumor therapy, and evade immunologic clearance (Kleffel and Schatton, 2013), indicating that there may be a direct relationship between DCCs and tumor dormancy. In other words, DCCs may be the main subtype of dormant tumor cells. Interestingly, besides DCCs-derived dormant tumor cells, non-DCCs-derived ones have also been discovered in many cancers, also suggesting that non-CSC subpopulations also have the capacity of becoming quiescent (Talukdar et al., 2019).

Disseminated tumor cells (DTCs) were first discovered in the bone marrow. These cells not only play a pivotal role in the growth, dormancy, early metastasis, and late recurrences of minimal residual disease (Goddard et al., 2018), but also mark the early stage of cancers and have been used to monitor tumor progression (Synnestvedt et al., 2012). To date, the specific source of DTCs is still unclear. However, recent studies proposed that DTCs could be derived from circulating tumor cells (CTCs) and CSCs at all stages of tumor progression (Crea et al., 2015; De Angelis et al., 2019). In addition, by analyzing cell phenotype and function, researchers have discovered that DTCs and CSCs both possess similar dormant stem-cell-like abilities, such as immune evasion, chemoresistance, and the ability to increase the likelihood of recurrence (Hen and Barkan, 2020). By definition, DCCs and DTCs, the progenitor cells of tumor dormancy, both have the potential ability to become dormant tumor cells (Jahanban-Esfahlan et al., 2019; Talukdar et al., 2019). For example, the latest research has found that the CSCs of breast cancer could lurk in the bloodstream and premetastatic sites as CTCs and DTCs during the early stage of the primary tumor for two decades (De Angelis et al., 2019).

Moreover, DTCs are involved in the formation of tumor dormancy mediated by the organ microenvironment and immune system (Sosa et al., 2014; Risson et al., 2020). For instance, the dormant DTCs can evade immune system recognition and seed in bone marrow or distant tissues as the root of future metastasis (Hosseini et al., 2016; Risson et al., 2020), suggesting that tumor metastases may occur during the early phases of tumor progression. In conclusion, these findings indicate that DTCs and DCCs, like many sleeping seeds, are undetectable but can switch between the state of sleeping and proliferation (Hen and Barkan, 2020). Besides, the cell intrinsic or microenvironmental factors can also determine whether a niche is permissive or not. The bone marrow, a kind of particularly permissive microenvironment for tumor cells, can maintain DCCs in a dormant state for many decades before reactivation and ultimate metastatic outgrowth (Hen and Barkan, 2020). To prevent cancer metastasis, in future studies, what initiates CTCs/DTC quiescence and what promotes them to re-enter the cell cycle need to be clarified.

Up to now, the mechanisms of tumor dormancy have only been partly illuminated, which can be divided into two categories: intracellular factors such as genetic alterations and the changes of signaling pathway; and extracellular factors such as tumor microenvironments, acidosis, hypoxia, angiogenic switch, as well as immunologic dormancy (Triana-Martinez et al., 2020). Therefore, only by deciphering mechanisms underlying MSC-mediated regulation of tumor dormancy, can more efficient targeted therapies be developed and the safe clinical application of MSCs be ensured.

Mesenchymal stem cells (MSCs) are a subset of stromal cells that possess the ability of multi-lineage differentiation and self-renewal (Ding et al., 2011). To be more specific, MSCs are reckoned as one subtype of adult stem cells, which not only have the capability of normal stem cells, but also show some unique multipotent properties (Kim and Kim, 2019) such as immune privilege, tumor homing feature, and anti-tumor ability (Eltoukhy et al., 2018b), making MSCs the most promising stem cells for curing cancer and ensuring safe clinical applications and trials. It has also been suggested that the effect of MSCs on tumor growth is contradictory, which can play anti-tumor or pro-tumor roles on tumor development (Bartosh et al., 2016). It can be concluded that MSCs are the most commonly used stromal cells in anticancer experimental or clinical researches because of their great therapeutic potential.

Current studies have shown that MSCs could interact with tumor cells and their niches through the paracrine pathway and physical mechanisms (Cammarota and Laukkanen, 2016), to mediate the growth of a primary tumor and the spread of distant metastasis (Ridge et al., 2017; Ahn, 2020). For example, the bone marrow-derived MSCs (BMSCs) can induce breast cancer cells into dormancy via the paracrine route to activate the JAK/STAT3 signaling pathway (Leyh et al., 2015), which partly illuminates the reasons why the metastatic latency period of breast cancer can prolong to 25 years after primary tumor resection. Studies have also found that whether MSCs promote or inhibit tumor progression and formation (Timaner et al., 2020) depends on the type of tumor, the source of MSCs, as well as tumor microenvironment. For example, relative studies have discovered that adipose tissue-derived MSCs (AMSCs) could promote the growth of brain tumors, ovarian tumors, gastric tumors, and breast tumor (Lim et al., 2016; Li et al., 2020), but inhibit the growth of melanoma both in vitro and in vivo (Ahn et al., 2015). Another study has proposed the allogeneic MSCs could induce melanoma cells into a malignant tumor (Djouad et al., 2003).

In 1986, Dvorak first proposed that the tumors are wounds that do not heal (Dvorak, 1986). In recent years, with the further research of cancer, it has been considered that MSCs are capable of regenerating and promoting wound healing (Lyu et al., 2020). However, tumor “wound” is different from conventional tissue wound, and MSCs can aggravate these non-healing wounds by directly or indirectly interacting with the tumor wound microenvironment (Li et al., 2019). For example, tumor-associated MSCs (TA-MSCs) usually promote conventional tissue wound healing via accelerating inflammation response and suppressing adaptive immunity. However, these two mechanisms are beneficial for tumor progression, resulting in cancer wound overhealing (Li et al., 2019). On the other hand, the bidirectional interaction between MSCs and tumor cells can result in constant stroma renewal, which is similar to a wound that never heals (Cammarota and Laukkanen, 2016).

In addition to the above studies, studies on the relationship between MSCs and tumor dormancy have also been extensively concluded in the antitumor field. During the migration into the tumor niche, MSCs can crosstalk with tumor cells and shape cancer phenotype (Spaeth et al., 2008). In this paper, the ways of different sources of MSCs influence tumor dormancy are summarized, providing more opportunities and challenges for researchers to focus on this field.

Mesenchymal stem cells have been discovered in most organs, including the menses blood, endometrial polyps, bone marrow, fetal tissues, umbilical cord matrix, and adipose tissue (Ding et al., 2011; Murphy et al., 2013; Li et al., 2020), but current studies have shown that different sources of MSCs possess distinct characteristics. These differences also partly illustrate why MSCs with different sources can play even opposite effects on tumor progression. It has been discussed in this paper that MSCs could either inhibit or promote tumor growth, but the specific mechanism is still controversial (Galderisi et al., 2010). Notably, tumor microenvironment plays an important role in regulating tumor dormancy, including the inhibition and promotion effects (Figure 1). For instance, TA-MSCs and BMSCs have different phenotypes, which may be influenced by tumor microenvironment factors, such as cancer cytokines and secretory protein (Cammarota and Laukkanen, 2016). In this review, the effects of BMSCs, hUCMSCs, TA-MSCs, and AMSCs on tumor dormancy are summarized. Table 1 shows the effect of MSCs on tumor dormancy.

Stem cell-based therapies have been used in anticancer researches for many years and MSCs are considered the most potent therapy tools. The reasons are summarized below. Firstly, MSCs can directly restrain the growth of some tumors by inhibiting the vasculature or arresting the cell cycle. Secondly, the engineered or modified MSCs have also been applied as targeted anticancer carriers for gene therapy (Li et al., 2020), which are more efficient and safer than naive MSCs. Moreover, MSCs can cross-talk with CSCs/DTCs via paracrine mechanisms, and regulate the biological activity of tumors (Lin et al., 2019), for example, AMSCs could suppress the growth of lung carcinoma cells by secreting cytokines such as interferon (IFN)-β and tumor necrosis factor-related apoptosis-inducing ligand (Cortes-Dericks and Galetta, 2019). In this review, the influence of MSCs on tumor dormancy is summarized and the ways of MSCs inducing CSCs and DTCs into dormancy is shown, helping to develop more promising treatment therapies.

During the past decade, researchers have realized that dormant tumor cells might be the primary reasons for therapy failure as the drug-resistant cells and dormant cancer cells share many similar biological features such as heterogeneity and plasticity (Luga et al., 2012), According to current clinical studies, there are two alternative strategies based on tumor dormancy. One way is to reawaken dormant tumor cells before treatment to make these cells susceptible to the therapies such as chemotherapy, immunotherapy, targeted therapy, and radiotherapy. For instance, one study has suggested that DTCs can be eliminated by manipulating their surrounding microenvironment to influence their communication with tumor stromal cells (Peinado et al., 2012). The other way is to drive dormant tumor cells into a perpetual dormant state. Because disseminated cancer cells could enter into the perivascular niche and remain dormant for decades, previous therapies can be applied to prolong the dormancy periods to prevent tumor metastasis and recurrence. For example, one clinical study has discovered the survival of patients with glioblastoma multiforme was significantly reduced, which may be caused by losing the control of dormant tumor cells. However, maintaining a perpetual tumor dormant state or preventing tumor cells from establishing dormancy can prolong the life of these patients (Ono et al., 2014).

More specifically, with the rapid development of stem cell-based therapies and molecular biology, great progress has been achieved in MSCs-based antitumor studies. It has shown that MSCs can be used to regulate tumor dormancy, which will provide more effective clinical strategies to prevent or delay cancer recurrence. Based on tumor dormancy, the potential clinical applications of MSCs on anticancer studies are summarized below.

Firstly, the MSCs and their exosomes can be used as carriers to target cancers. It is acknowledged that MSCs are capable of homing to sites of tissue injury and tumor microenvironment by reducing immune-inflammatory responses (Eltoukhy et al., 2018b). Based on this feature, researchers have proposed that MSCs could be used as carriers to deliver nanoparticles, anti-tumor drugs, proteins, lipids, DNAs, mRNAs and miRNA (Eltoukhy et al., 2018b; Kwon et al., 2019). For example, one study reported that internalization of paclitaxel-loaded nanoparticles to MSCs can achieve promising antitumor efficacy, which is caused by sustained release of the encapsulated paclitaxel to the tumor microenvironment (Wang et al., 2018). Another study proposed that BM-MSCs can induce breast CSCs into dormancy through releasing inhibitory miRNA such as antagomiR-222/223 (Ono et al., 2014; Eltoukhy et al., 2018b). Moreover, Lai et al. have conducted more detailed studies and discovered that the exosomes derived from MSCs are better candidates for drug delivery due to natural features such as easy isolation, strong tolerability, as well as the ability to bind with the plasma membranes (Vallabhaneni et al., 2017). It is believed that MSCs–derived exosomes can be used to prevent the recurrence of cancer. Although in the experimental investigation stage, these studies have developed a new therapeutic strategy for cancer.

Secondly, cannibalism of MSCs can be used to assess the prognosis of cancer. Current studies have suggested that the cannibalism of MSCs can promote tumor dormancy, and happen in highly aggressive tumors most frequently (Bartosh et al., 2016). Therefore, Thomas et al. have proposed that there is a logical cause-and-effect relationship between cell cannibalism and dormancy (Bartosh et al., 2016). To be more specific, cannibalistic cells can be used to identify cancer phenotype, which will provide new windows for cancer therapeutic intervention and prognostic evaluation. Furthermore, with the improvement of the 3D coculture model, more mechanisms underlying the mediation of cell cannibalism by MSCs will be unraveled and more antitumor tools can be used in clinical application.

Third, the formation of tumor dormancy can be intervened by targeting the key players of interaction between MSCs and tumor cells. Until now, with the discovery of more and more key players regulating organ-specific metastasis and dormancy (Neophytou et al., 2019), targeted therapy has been frequently used in clinical applications. Directly targeting tumor cells is highly effective cancer therapy. However, the MSCs can interact with CSCs and drive them into a state of cycling dormancy through secreting proteins, exosomes, and cytokines, which will make these CSCs resist targeted therapy (Lim et al., 2011; Patel et al., 2012; Ono et al., 2014). Therefore, to make these CSCs drug-sensitive, the interaction between the CSCs and MSCs needs to be disrupted (Eltoukhy et al., 2018a). One of the most promising strategies for reducing bone metastasis is to target dormancy-associated niches. In clinical practice, bone metastasis is reckoned as an indicator of poor prognosis in most cases. With the description of the “bone metastatic niche” concept, which consists of connective tissues, bone stromal cells, and signaling molecules, some researchers have proposed that bone metastatic niche can be targeted to prevent or delay the progression of bone metastasis (Ren et al., 2015). For example, BMSCs, one subtype of bone stromal cells, can either promote or impede the progression of tumor dormancy by secreting BMP7, TGFβ2, GAS6 (Kaplan et al., 2006; Decker et al., 2017; Hu et al., 2020; Li et al., 2020). Targeting the bone metastatic niche can disrupt the interaction between MSCs and tumor cells, thus restraining the formation of tumor dormancy. Consistent with these discoveries, another study has suggested that dormant tumor cells could become sensitive to normal cytotoxic therapies by targeting at DTC dormant niche (Saito et al., 2010). All in all, directly targeting the mechanisms underlying tumor dormancy is a new clinical cancer strategy.

Finally, some researchers have proposed that immune therapy combined with dormancy therapy would develop a new cancer therapy (Linde et al., 2016). For example, immunotherapy has been successfully used in maintaining the dormancy of HIV, which has many similarities to tumor cells (Boire et al., 2019). Simultaneously, MSCs can either suppress or enhance the immune function of cancer cells as influencing tumor proliferation (Roccaro et al., 2013). In conclusion, a better understanding of the immunobiological pathways of dormant tumor cells will help study the mechanisms of how MSCs mediate tumor dormancy (Kleffel and Schatton, 2013).

Tumor dormancy has attracted more and more attention from researchers. However, related investigations are limited by the lack of classical models that can represent the human condition. The essential effects of MSCs on tumor dormancy have been discussed. The influence of bone marrow microenvironment such as fibroblasts, endothelial cells, and immune cells can also not be ignored (Reddy et al., 2012). However, it is difficult to recreate the structure of the bone marrow, limiting the studies of other cells within the bone marrow. In the past several years, many researchers have used the 2D model of the bone marrow to replicate the interaction between cellular and soluble factors (Chong Seow Khoon, 2015). However, as time goes on, traditional 2D co-cultures platforms have shown deficiencies (Widner et al., 2018). For instance, 2D co-cultures could not adequately mirror the natural bone marrow microenvironment and physiological conditions of cells, which greatly limits the therapeutic testing and system modeling in vitro (Bartosh et al., 2016).

To better study the interactions between bone marrow and tumor dormancy, it is essential to model a platform microenvironment where different types of cells can be co-cultured and cell phenotype can be analyzed (Eltoukhy et al., 2018b). Fortunately, in recent years, with the emergence of 3D bioprinting and engineered organotypic models, an increasing number of fundamental problems of tumor biology have been addressed. For example, an engineered organotypic microvascular niche was utilized to study the interaction between endothelial cells and breast cancer cells. The results showed that endothelial-derived thrombospondin-1 could directly induce sustained breast cancer cells quiescence (Ghajar et al., 2013). Similarly, a 3D system can be created to recapitulate the bone microenvironment (Gungor-Ozkerim et al., 2018). This system can be used to better understand the biological properties of MSCs and mirror natural conditions in vivo (Clark et al., 2018). Besides, the model can be used to create an MSCs-tumor stem cell co-culture system, which greatly helps us investigate the interaction between the tumor cells and their niches (Sart et al., 2014; Moore et al., 2018). Likewise, we will easily select cell types and figure out which cells and niches support tumor dormancy (Eltoukhy et al., 2018b). However, though these models have made a great contribution, the whole picture of dormancy-related cellular processes is still far from clear (Cole, 2009). Undoubtedly, only when the data collected from different dormancy models are analyzed together, may the stealth mechanisms of tumor dormancy start to be clarified.

In recent years, a plenty of mouse models have been established to reveal the mechanisms of various diseases. However, it is still difficult to validate the molecular mechanisms of dormancy in mouse models. First of all, due to the fact that tumor dormancy is a phenomenon that occurs at low frequency, developing a well-established mouse model of tumor dormancy is probably impractical. In addition, each mechanism of tumor dormancy has different complex characteristics, and current mouse models cannot precisely monitor and manipulate single or small clusters of tumor cells. On the other hand, even there are applicable conceptual models of dormancy, the result of these experimental studies is still difficult to be validated in patients. Therefore, it is very important to establish a sophisticated experimental dormant model and to isolate CSCs and DTCs from the tumor microenvironment, and then simulate the interaction between dormant tumor cells and the surrounding niches. Currently, since the dormant tumor cells are undetectable via traditional whole-body imaging tools (Linde et al., 2016), it is better to develop new diagnostic tools that can detect dormant tumor cells in patients as soon as possible. In the long run, if more biomarkers of dormant tumor cells can be identified before tumorigenesis or more quantitative methods and calculable tools can be used in clinical diagnosis, the recurrence and metastasis of cancer will be efficiently prevented.

Whether MSCs promote or suppress tumor dormancy is still controversial. From an individual point of view, based on their sources, MSCs can be divided into two types: tumor-derived MSCs and other sources-derived MSCs. The tumor-derived MSCs are progeny of tumor cells, which mainly play a pro-tumor role during the tumor progression but other sources-derived MSCs play an anti-tumor role. However, this rule is not absolute and depends on the types of tumors. Aparts from being influenced by the sources of MSCs and the types of cancer, the effects of MSCs on tumor progression can also be influenced by the experimental models. Therefore, the interpretation of some experimental results was somewhat inconsistent, even the opposite. In the future, with the development of the 3D bioprinting technique, the modeling methods will be greatly improved.

In recent years, MSC-based therapies have been frequently used by clinical researchers. During these processes, MSCs are modified through the transfection of therapy genes and transportation of target agents, but the inherent characteristics and biological properties of MSCs are not changed. Combining with the studies of tumor dormancy, researchers can develop more modified MSCs to intervene in the progression of tumor dormancy. Although more detailed mechanisms underlying the interaction between MSCs and tumor cells needs further study, current studies are trying to apply engineered MSCs as treatment carriers to regulate tumor dormancy to prevent or delay tumor progression. All in all, the treatment potential of MSCs on tumors is infinite. Although it is difficult to apply these strategies clinically due to the limitations of current technical merit, there is no doubt that a huge step forward has been taken in the fight against cancer.

LZ and KZ drafted the review. HH generated the graphs. YY guided the construction of the manuscript. WL edited the review. JL and TL provided input on the scope and content of the review. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.