Colorectal cancer and dormant metastases: Put to sleep or destroy?

Abstract

After reading the review by An et al “Biological factors driving colorectal cancer metastasis”, which covers the problem of the metastasis of colorectal cancer (CRC), I had a desire to discuss with readers one of the exciting problems associated with dormant metastases. Most deaths from CRCs are caused by metastases, which can be detected both at diagnosis of the primary tumor and several years or even decades after treatment. This is because tumor cells that enter the bloodstream can be destroyed by the immune system, cause metastatic growth, or remain dormant for a long time. Dormant tumor cells may not manifest themselves throughout a person’s life or, after some time and under appropriate conditions, may give rise to the growth of metastases. In this editorial, we will discuss the most important features of dormant metastases and the mechanisms of premetastatic niche formation, as well as factors that contribute to the activation of dormant metastases in CRCs. We will pay special attention to the possible mechanisms involved in the formation of circulating tumor cell complexes and the choice of therapeutic strategies that promote the dormancy or destruction of tumor cells in CRCs.

Key Words: Circulating tumor cells, Colorectal cancer, Disseminated tumor cells, Dormant metastases, Metastases, Premetastatic niche

Core Tip: After reading the review by An et al “Biological factors driving colorectal cancer metastasis” (World J Gastrointest Oncol 2024, 16: 259-272), I had a desire to discuss with readers the problem associated with colorectal cancer (CRC) dormant metastases. Metastases can be detected during the initial diagnosis of CRC or can appear many years after treatment. Late metastases are caused by dormant tumor cells. This editorial discusses the most important features of dormant metastases and premetastatic niches, factors that promote the activation of dormant metastases, and possible therapeutic strategies to promote tumor cell dormancy or their killing in CRCs.

INTRODUCTION

Despite the fact that, owing to effective screening, mortality from colorectal cancer (CRC) is decreasing in most developed countries, CRC remains the second most common cause of cancer-related death worldwide[1]. Most deaths from CRCs, as well as from other malignant neoplasms, are caused by metastases[2,3]. The survival rate of patients with metastatic CRC does not exceed 15%[2]. CRC metastases can be detected both at initial diagnosis (in approximately 20% of patients with CRCs) and years after tumor treatment. It is believed that more than 30% of patients with early CRCs will subsequently develop metastases[4]. CRCs most often metastasize to the liver (approximately 70% of patients), followed by metastases to the lungs, distant lymph nodes and peritoneum[5,6]. An interesting feature of CRCs is the relatively frequent formation of single metastases in the liver and lungs, the combined treatment of which (surgery and systemic therapy) can lead to long-term remission[6,7].

Numerous studies have identified the factors contributing to cancer metastasis and established the main stages of the metastatic cascade. A review by An et al[3] titled “Biological factors driving colorectal cancer metastasis” (World J Gastrointest Oncol 2024, 16: 259-272) focused on the biological factors that drive CRC metastasis, namely, driver mutations, genomic instability and epigenetic instability. The authors discussed in detail the role of epithelial-mesenchymal transition (EMT), angiogenesis, and hypoxia in CRC metastasis and pointed to the involvement of exosomes in transducing key signals promoting CRC progression. The authors paid special attention to the influence of the innate and adaptive immune systems; lifestyle factors, such as overweight/obesity, physical inactivity, cigarette smoking, alcohol consumption and inappropriate dietary patterns on the metabolism of tumor cells and mechanisms of tumor progression in patients with CRCs[3]. However, it should be noted that most studies have not revealed fundamental differences in the frequency of mutations in early CRC cells and metastatic cells. In CRCs, only TP53 mutations are notably more frequent in metastatic cells than in early CRC cells[3]. Moreover, comparison of mutational profiles of primary CRCs and their matching ovarian metastases suggested that dormant metastases may be present from the early days of primary tumor formation[8]. These data indicate that our knowledge is still limited by the incomplete understanding of the mechanisms of CRC metastasis.

In this editorial, I would like to discuss one of the most interesting and intriguing problems in oncology, the problem of dormant metastases, with readers. Currently, it has been established that tumor cells that enter the bloodstream have at least three possible fates. These cells can be destroyed by the immune system, cause metastatic growth to occur, or remain dormant for a long time, leading to dormant metastases. Dormant tumor cells may not manifest themselves throughout a person’s life or, after some time and under appropriate conditions, may give rise to the growth of metastases. One of these conditions necessary both for maintaining the dormant phenotype of tumor cells and for “rooting” and growing metastases is the presence of a suitable environment or “premetastatic niches”[9,10].

The problem of dormant metastases is definitely so interesting and multifaceted that increasingly more new researchers are turning their attention to it. This interest is associated not only with the mysterious state of tumor cells (these cells can remain silent for years or even decades) but also with promising opportunities for establishing control over the course of the disease, that is, for achieving controlled remission. In this editorial, we will discuss some of the most important features of dormant metastases and premetastatic niches (PMNs), factors that promote the activation of dormant metastases, and possible therapeutic strategies to promote the dormancy or eradication of tumor cells in CRC.

WHAT ARE DORMANT METASTASES?

Previously, it was assumed that tumor progression is associated with the continuous accumulation of somatic and epigenetic mutations from the time of cancer initiation to the time of metastasis[11,12]. However, cancer metastases can occur several years or even decades after successful resection of the primary tumor. The appearance of late metastases is explained by the ability of disseminated tumor cells (DTCs), which survive anticancer treatment, to remain dormant for a long time[13-15].

In scientific publications devoted to the mechanisms of cancer metastasis, two terms are often used: Circulating tumor cells (CTCs) and DTCs. Although all DTCs originate from CTCs, the two terms are similar but not identical[14,15]. CTCs are cells that are found in the blood of cancer patients. The results of clinical studies indicate that the number and characteristics of CTCs are closely related to the prognosis and effectiveness of drug therapy[16]. In CRC, the total number and phenotype of CTCs are effective markers for assessing disease prognosis and treatment efficacy[17,18]. In particular, positive correlations between the number of CTCs and TNM stage, T stage, N stage, M stage and CRC prognosis have been established[19]. Several studies have shown that, compared with CTCs with an epithelial phenotype, CTCs with a mesenchymal or hybrid epithelial-mesenchymal phenotype are more strongly associated with decreased overall and disease-free survival[19,20]. Moreover, different CTC phenotypes were noted for right-sided and left-sided colon tumors. In right-sided CRC, CTCs exhibit a predominant apoptotic pattern, while in left-sided colon cancer, a mesenchymal phenotype predominates[21]. The authors believe that these data indicate differences in the biology of cancer in the right and left halves of the colon and differences in the fate of CTCs during these localizations of CRC.

Despite the importance of studying CTCs, these cells are unable to form metastases until they reach their target organs. DTCs are tumor cells that are found at distant sites, such as the bone marrow, lymph nodes, liver, lungs, and other sites. DTCs are found both in patients with metastases and in cancer patients at the earliest stages of the disease[12]. The most acceptable way to identify these lesions is via bone marrow examination. However, this method is very labor intensive. In addition, the absence of DCM in the bone marrow does not guarantee its absence in other organs, such as the lungs or liver.

As mentioned above, all DTCs originate from CTCs. However, the fates of CTCs can differ: They can be destroyed by the immune system or drugs, can turn into DTCs and initiate the growth of a metastatic site, or can turn into dormant metastases. Dormant metastases are defined as cancer cells that have entered prolonged growth arrest (G0/G1), despite the presence of promitogenic and survival-promoting mutations[22,23]. Numerous studies have allowed us to identify some features of dormant tumor cells. In particular, dormant tumor cells are characterized by high expression of nuclear receptor subfamily 2, group F, member 1 (NR2F1) and have characteristics of both embryonic and adult cells[24-26]. In an experiment, NR2F1 activated the resting and self-renewal genes SOX9, OCT4, SOX2, and NANOG. However, at the same time, high expression of NR2F1 negatively correlated with Ki67 expression[25]. In CRC, activation of NR2F1 was also associated with a decrease in the proliferation rate of tumor cells[27,28].

The internal signaling pathways associated with the transition to a dormant state in cancer cells include a decrease in the activity of extracellular regulated kinase (ERK1/2) and activation of p38 mitogen-activated protein kinase (p38 MAPK). The balance between ERK1/2 and p38 MAPK signaling is regulated by fibropectin and urokinase-type plasminogen activator (uPA) signaling through the uPA receptor and specific integrins[26,29,30].

The transition of DTCs to a dormant or proliferating state is determined by the existing microenvironment, which differs among target organs. For example, DTC dormancy can be induced by TGFβ2 through p38-dependent signaling. p38 MAPK blocks the cell cycle and induces a dormant state by suppressing cyclins and increasing the activity of cyclin-dependent kinase inhibitors (such as p21, p27, and p16) through the activation of NR2F1[24,26,29,31]. DTC dormancy can also be induced by bone morphogenetic protein 7 (BMP7), which is of bone marrow origin[32], and osteoblasts[33]. In addition, thrombospondin-1, produced by endothelial cells, can promote the transition of DTCs to a dormant state[34].

The data on the influence of TGFβ2 and BMP7 on the process of CRC metastasis are notably contradictory. Analysis of TGFβ2 expression in CRC tissues revealed that the expression of this marker was greater in tumors than in normal tissues[35]. Using the Kaplan-Meier plotter database to assess the relationship between transforming growth factor beta 2 (TGF-β2) expression and CRC prognosis, the authors found that a high level of TGFβ2 was associated with a better prognosis in CRC patients than a low level of this marker. However, according to the GEPIA database, high TGF-β2 expression is significantly correlated with poor prognosis in CRC patients[35]. In an experimental study by Tauriello et al[36], TGFβ inhibition induced a potent and durable cytotoxic T-cell response that prevented CRC metastasis. Moreover, in mice with advanced metastatic liver disease, blockade of TGFβ signaling rendered tumors susceptible to anti-PD-1/PD-L1 therapy.

Studies of the role of BMPs in CRC progression have shown that BMP signaling depends on the CRC subtype. In the mesenchymal subtype of CRC, high BMP activity promoted synergistic interactions with BMP-Notch, which correlated with decreased survival in CRC patients[37]. However, in an experimental study by Karagiannis et al[38], BMP signaling, in contrast, reduced the proliferation and invasion of CRC cells and disrupted EMT. The inconsistency of the results obtained may indicate that the same factors, depending on the characteristics of the tumor cells and the microenvironment, can have different effects on DTCs, either by inducing a dormant state or activating them. For example, the immune system can both promote the transition of DTCs to a dormant state and, on the contrary, activate dormant tumor cells[31]. The role of the immune system will be discussed in more detail in the next section.

The transition of DTCs to a dormant state can be influenced by the hypoxic microenvironment by enhancing the transcription of genes associated with EMT [such as SNAIL, zinc finger E-box-binding homeobox 1 (ZEB1), TWIST and TCF3] and the stem-like phenotype of tumor cells[39]. For this reason, dormant tumor cells exhibit both stem-like and EMT phenotypes, as indicated by the expression of CD44, vimentin, snail, CD133, and SOX2[31,40,41]. In addition, dormant tumor cells exhibit low expression of markers associated with tumor cell proliferative activity (e.g., Ki67, c-Myc, and cyclin D1)[42].

The transition of tumor cells to a dormant state is also associated with the activation of autophagy in response to stress. Autophagy plays an important role in the adaptation, survival and reactivation of dormant cells[43]. The use of autophagy inhibitors, such as hydroxychloroquine or CDK4/6 inhibitors, can promote the awakening of dormant tumor cells and their subsequent destruction by chemical or targeted drugs[22].

DTCs are often detected in the bone marrow of patients with various malignant neoplasms. Factors that promote the dormancy of hematopoietic stem cells in the bone marrow are hypothesized to also cause dormancy in DTCs[26,44]. This hypothesis suggests the existence of a specialized microenvironment in perivascular niches that can promote survival and impose growth restrictions on both stem cells and DTCs. Moreover, it is possible that the number of dormant niches is limited, and as the primary tumor grows and the number of CTCs increases, the latter, entering already occupied niches, will activate the tumor cells located there, promoting the formation of metastases. This assumption is supported by the fact that the same cues used to mobilize hematopoietic stem cells from bone marrow also result in the activation of dormant tumor cells[44].

Antitumor drug therapy can also promote the transition of tumor cells to a dormant state[23,45]. RNA-seq of dormant ovarian cancer cells immobilized in solid silica gel revealed that genes associated with platinum resistance pathways were expressed. These cells exhibit increased viability and retain the ability to proliferate. Moreover, despite their ability to proliferate, these cells exhibit increased resistance to cisplatin and paclitaxel[46]. However, it is possible that antitumor therapy promotes the selection of cells that are resistant to treatment, and their transition to a dormant state is determined by the volume of the residual tumor and the microenvironment. This is evidenced by the fact that in patients with complete therapeutic pathomorphosis of the tumor and metastases (RCB0), the prognosis is better than that in patients with a large residual tumor (RCB3-4)[47,48].

PREMETASTATIC NICHES CAN APPEAR LONG BEFORE METASTASIS APPEARS

As already noted, the fates of CTCs and DTCs can differ and are determined both by the characteristics of the tumor cells themselves and by their microenvironment in special niches located in target organs, called PMNs. For tumor cells to enter PMNs, their exit from blood vessels or extravasation is necessary. The process of extravasation occurs with the participation of adhesion molecules such as integrins, the transmembrane protein mucin-1 (MUC-1) and CD44. In metastatic CRCs and breast cancers, the adhesion of tumor cells to the adjacent endothelium is achieved through the interaction of MUC-1 with intercellular cell adhesion molecule-1, E-selectin, and galectin-3[49].

The formation of PMNs is believed to involve a series of sequential events, such as the formation of obstacles for tumor cells in the microvasculature and impaired vascular permeability; the modification of the extracellular matrix; the cellular reprogramming of immune cells; and the activation of proinflammatory molecules such as S100, tumor necrosis factor alpha, and TGF-β[50-52]. These changes may occur long before tumor cells arrive at PMNs. On the one hand, primary tumors and DTCs can influence the formation of premetastatic/perivascular niches through exosomes[53,54]. It has been found that tumor-derived exosomes can promote microvascular thrombosis, thereby increasing the risk of metastasis[50]. In addition, exosomes can support cancer cell proliferation and metastasis by stimulating MAPK and PI3K-Akt signaling, enhancing matrix metalloproteinase expression, and promoting cancer cell adhesion to the endothelium and fibrinogen[55]. The immunosuppressive effect of exosomes manifests as the accumulation of myeloid-derived suppressor cells in PMNs, the suppression of T cells and NK cells, the impairment of dendritic cell maturation in lymph nodes and increased levels of circulating exosomal PD-L1[52,56-58].

In an experimental CRC model, tumor-associated macrophages stimulated tumor cells to produce CXCL1, which led to an increase in its amount in the liver. In liver tissue, CXCL1 promoted the recruitment of CXCR2-positive myeloid suppressor cells (MDSCs) to PMNs, which led to the formation of liver metastases[59]. In another experimental model of CRC, primary tumors released integrin beta-like 1-rich extracellular vesicles, which activated resident fibroblasts in distant organs. Activated fibroblasts induced the formation of PMNs and promoted the growth of metastases through the secretion of proinflammatory cytokines such as interleukin (IL)-6 and IL-8[53]. Thus, the immune microenvironment of PMNs may influence DTCs, promoting their transition to a dormant state, or, conversely, may contribute to their awakening[10,50]. In addition, the perivascular niche may contribute to therapeutic resistance and tumor cell evasion from the immune system[51].

Interesting results were obtained by Ren et al[42]. The authors observed the development of liver metastases in mice with and without resection of the primary tumor in an experimental model of CRC[42]. In mice with primary tumor resection, micro- and macrometastases developed at 4 wk and 4 months, respectively, whereas in mice without primary tumor resection, micro- and macrometastases developed at 2 wk and 8 wk, respectively. In contrast to CRC cells in macrometastases, tumor cells in micrometastases exhibited characteristics of dormant tumor cells, as demonstrated by increased expression of CD44, vimentin, and CD133 and decreased expression of Ki67, c-Myc, and E-cadherin. The immune profiles of micro- and macrometastases also differed. Liver tissue with micrometastases had more dendritic cells, CD8+ T lymphocytes, and macrophages, whereas liver tissue with macrometastases accumulated MDSCs. In vitro, cocultivation of MDSCs with dormant tumor cells promoted the release of the latter from the dormant state, which was manifested by decreases in the levels of CD133 and SOX2 and increases in the levels of cyclin D1 and c-Myc. Coculture of MDSCs with CD8+ T lymphocytes reduced the number of IFN-γ-positive cells. The authors showed that chemokine (C-C motif) ligand 7 (CCL7), synthesized by MDSCs, plays a crucial role in the activation of dormant CRC cells. CCL7 binding to CCR2 in dormant tumor cells activated the JAK-STAT3 pathway. A CCL7 inhibitor prevented the development of CRC metastases and prolonged the survival of mice. In addition, in CRC patients with liver metastases, the number of MDSCs was greater, and the serum CCL7 level was greater than that in patients without metastases[42]. These data indicate the importance of early CRC diagnosis and the role of immune surveillance in maintaining the dormant state of DTCs.

Neutrophil extracellular traps (NETs) play important roles in all stages of metastasis formation in PMNs, including intravasation, circulation of tumor cells in the bloodstream, and extravasation and activation of dormant metastases. NETs are web-like fibers containing extracellular DNA, myeloperoxidase, and proteins released by neutrophil degranulation[60]. NETs are activated by CTCs and participate in their spread and activation of metastatic growth in target organs[61,62]. By increasing vascular permeability, NETs promote the extravasation of tumor cells[63] and increase the risk of metastasis associated with surgery[64], including in CRCs[65]. Several experimental studies have shown the active formation of NETs in target organs of CRCs under the influence of various factors, for example, in the liver during systemic infection[66] or in the lungs during chronic inflammation[67]. It has been suggested that neutrophils and NETs may be involved in both the creation of PMNs and the awakening of dormant tumor cells, causing disease relapse and cancer metastasis in the long term after treatment of the primary tumor[62,66,67].

In CRC, NETs were identified in the primary tumor and draining lymph nodes with metastases[68]. In an experiment, exosomal KRAS mutation promoted the growth of CRC cells by increasing the production of IL-8 and NETs[69]. IL-8 is known to be a chemotactic stimulus that promotes tumor cell migration and angiogenesis. In addition, IL-8 can influence the proliferation and survival of cancer cells by promoting invasion, growth and metastasis[70]. IL-8, which interacts with its receptor CXCR2 on neutrophils, induces neutrophils to release NETs by activating Src, ERK, and p38 signaling. In turn, NETs can activate nuclear TLR9, promoting cancer progression[71]. Associations between NETs, interleukins, and Toll receptors and the risk of CRC metastasis have been demonstrated in in vitro, in vivo and clinical studies[64,72,73]. In CRC, increased numbers of neutrophils in the tumor stroma and PMNs may be directly related to the translocation of intestinal bacteria involved in CRC carcinogenesis[74,75].

However, it should be noted that in a number of experimental studies, NETs prevented the growth of cancer cell cultures by inducing apoptosis and/or inhibiting proliferation[68], which may have been associated with the absence of CXCR2 (the IL-8 receptor) on tumor cells[76]. Thus, the effect of NETs on tumor cells may depend on various factors that have yet to be established.

Another factor that may influence the fate of DTCs in PMNs is the stiffness of the extracellular matrix. In an experiment, cancer cells cultured on hard supports exhibited a proliferative phenotype due to activation of β1 integrin-mediated ERK, Akt, and STAT3 signaling, whereas cancer cells cultured on soft supports exhibited dormant phenotypes with increased expression of stemness-related markers and resistance to therapy[77,78]. Similar results were obtained by other authors. Wang et al[79], in an experimental model of metastatic liver damage, observed the transition to a dormant state of tumor cells in the elastic framework of liver tissue. However, in the rigid framework of liver tissue, tumor growth created pressure within the tumor, which led to tissue stretching in the nearby liver parenchyma. Apoptosis and death of tumor and liver cells as a result of mechanical action led to activation of tumor cell proliferation and the growth of metastatic foci along the paths of least mechanical resistance[79]. In humans, changes in liver tissue stiffness can be caused by various processes, such as fibrosis, injury, and aging.

Notably, the recognition of PMNs provides new opportunities for the prevention of tumor cell metastasis[10]. However, their fundamental characteristics are still being discussed, and new experimental and clinical studies are needed to develop new effective methods for the diagnosis, treatment and prevention of malignancies.

WHAT CAN AWAKEN DORMANT METASTASES AND WHAT ARE POSSIBLE THERAPEUTIC STRATEGIES?

The formation of metastases and the awakening of dormant tumor cells are determined both by the characteristics of the primary tumor (tumor grade, heterogeneity, number and nature of accumulated mutations) and by the nature of the microenvironment in PMNs, where tumor cells enter and especially in niches around the microvasculature[10]. Factors such as aging, inflammation, hypoxia, hormonal changes, and lifestyle can directly impact the behavior of DTCs, stimulating their growth and the formation of metastases[10,79,80]. In contrast, epigenetic drugs such as azacitidine and retinoic acid, as well as small molecule NR2F1 agonists, can induce DTC dormancy, inhibiting the formation of metastases[10].

We next considered several factors that may be associated with the activation of dormant tumor cells and the formation of metastases.

Hypoxia

Hypoxia leads to the activation of key genes associated with hypoxia (e.g., GLUT1 and HIF1α), EMT and the tumor cell stemness phenotype (e.g., Snail, TWIST, and Notch), as well as the activation of genes associated with the dormant state of tumor cells (e.g., NR2F1, DEC2, and p27)[39,81,82]. However, global changes in the levels of reactive oxygen species (ROS) can lead not only to a dormant state of tumor cells but also to their active growth[83]. An increase in ROS may also be associated with infection, stress, chemotherapy and radiation therapy[81]. Given these data, the use of antioxidant drugs may help maintain the dormant state of tumor cells[84]. In contrast, the combined use of drugs that promote the activation of dormant metastases (e.g., interferon alpha) and anticancer drugs can aid in the eradication of dormant tumor cells[83].

Intracellular bacteria

Increasing evidence suggests the presence of intracellular bacteria not only in primary tumors but also in metastases[85]. In CRC, Fusobacterium nucleatum (F. nucleatum) was detected in 82% of primary tumors and in 64% of CRC liver metastases. Moreover, 99.9% of the bacterial strains identified in primary tumors and metastases had similar nucleotide sequences, even if the tissues from the primary tumor and metastases were collected several months or even years apart[74]. Moreover, negative correlations were noted between the abundance of F. nucleatum and the survival of patients with CRC, indicating a possible connection between the intracellular persistence of bacteria and the progression of CRC. In a study by Bullman et al[74], CRC tumors positive for Fusobacterium were successful xenografts, whereas Fusobacterium-negative tumors could not be cultured. Oral administration of metronidazole to mice bearing Fusobacterium-positive xenografts resulted in decreased tumor growth and tumor cell proliferation[74].

A study by Bertocchi et al[75] revealed that the intestinal microbiota, by disrupting the intestinal vascular barrier, promotes the dissemination of bacteria into the liver, thereby influencing the formation of PMNs and promoting the metastasis of CRCs[75]. In patients with CRC, increased levels of PV-1, a marker of intestinal vascular barrier disruption, were associated with bacterial dissemination in the liver, metachronous distant metastases and decreased 10-year disease-free survival (40% vs 72% for patients with high PV-1 and low PV-1, respectively; P < 0.0001)[75]. Thus, it can be assumed that the persistence of bacteria may be one of the factors contributing to the awakening of dormant metastases due to the influence on the cytoskeleton of tumor cells, participation in the metabolism of chemotherapeutic drugs, modulation of the immune response, destruction of the vascular barrier and induction of inflammatory changes at distant sites[85,86]. The use of antibacterial drugs against pro-carcinogenic intestinal bacteria and normalization of the intestinal microbiota may improve long-term outcomes of CRC treatment, as evidenced by the results of several studies[87].

NETs

In CRC, the mechanism of NET formation may be associated with procarcinogenic bacteria, such as F. nucleatum[88]. CRC patients infected with F. nucleatum had greater tumor infiltration by neutrophils and greater levels of NETs in tumor and blood samples. In vitro, NETs induced by F. nucleatum accelerated tumor growth through angiogenesis and promoted metastasis associated with EMT. NET formation was associated with the activation of TLR4-ROS signaling and the NOD1/2 receptor[88]. In an experiment, treatment with DNase, neutrophil elastase inhibitor or carcinoembryonic Ag cell adhesion molecule 1 (CEACAM1) reduced the metastasis of several cancers[66], including CRC[72,73,89]. Similarly, blocking the IL8-CXCR2 axis or inhibiting TLR9 slowed tumor progression in preclinical models[71]. It can be hypothesized that this therapy may be effective at preventing surgery-associated CRC metastasis. However, the possibility of its use for the prevention of late metastases of CRC is unclear.

The role of CTC clusters and heterogeneity of the primary tumor and metastases

It has recently been established that CTCs can unite in clusters and that these clusters have a greater ability to form metastases than single tumor cells. Complexes of CTCs with platelets, myeloid cells or tumor-associated fibroblasts have similar properties[90-93]. Indeed, in patients with metastatic CRC, both single CTCs and clusters of CTCs were detected in the blood. The presence of CTC clusters has been associated with increased levels of TGF-β and CXCL1 and decreased overall survival[94]. In an experimental CRC model, clusters of tumor cells orthotopically injected into the submucosa of the rectum of NOD mice formed liver metastases more often than single tumor cells. Circulating clusters of CRC cells contained hybrid tumor cells expressing E-cadherin and ZEB1. Inhibition of these factors reduced metastasis to the liver in mice[95]. In vitro, clusters of CRC cells expressed higher levels of genes responsible for cancer stemness (CD133 and Lgr5), EMT (E-cadherin and TGF-β 1-3), hypoxia, and CRC surface markers (including CD24, CD44, and CD133) than did individual CTCs. The authors believe that the main factor promoting the binding of cells to clusters may be CD24[96].

Notably, we previously described a possible mechanism for the formation of tumor cell clusters in tumor microvessels. We hypothesized that as a result of disruption of the adhesive properties of tumor cells, they may detach from the underlying substrate, which leads to the formation of “hollow” structures with tumor cells in the lumen. The described structures can then be lined by endothelial cells and subsequently merge with blood or lymphatic vessels[97]. We call this type of tumor microvessel formation the “cavitary” type of tumor angiogenesis. Its markers include the phenomenon of peritumoral retraction clefting and the presence of structures with partial endothelial lining. In gastric cancer, breast cancer and cervical cancer, the presence of these phenomena in the tumor stroma was associated with a high risk of regional metastases and disease relapse[97-99].

The more aggressive behavior of tumor cell clusters may be explained by the fact that large clusters of CTCs are more easily attached to the microvessels of target organs, and the hypoxic environment ensures the activation of transcription factors associated with stemness and proliferation, such as OCT4, NANOG, SOX2, and SIN3A[100]. Сirculating clusters of tumor cells retain the ability to pass through the capillaries of target organs by unfolding into single-file chains[101]. In addition, large tumor clusters are more likely to contain heterogeneous cell types than small clusters or single tumor cells.

The role of heterogeneity of the primary tumor and CTCs in the progression of malignant neoplasms has been noted by many authors[95,102-104]. In patients with prostate cancer in remission, DTCs in the bone marrow had a dormant transcriptome with a high content of NR2F1/p38K, while in patients with disease recurrence, both dormant DTCs and proliferating DTCs were found[26,105]. Primary breast cancer tumors with regional and distant metastases were characterized by intratumoral heterogeneity with low phosphoglycerate dehydrogenase (PHGDH) expression, in contrast to breast cancer without metastases, where intratumoral homogeneity with high expression of PHGDH was observed[106]. In an experimental study by Heinz et al[102], the formation of CRC macrometastases was observed only in samples containing heterogeneous types of tumor cells, both Lgr5- and Lgr5+. In contrast, micrometastases contained only Lgr5- cells, which were characterized by a lack of growth; that is, they showed signs of dormant metastases. Thus, the heterogeneity of DTC complexes, in which stem and differentiated cells are present, appears to be one of the conditions for activating the growth of dormant metastases.

Surgical intervention

It has now been established that surgery may be a risk factor for the activation of dormant micrometastases and the acceleration of the appearance of new metastases, including in CRC[107,108]. Removal of the primary tumor may allow malignant cells to escape into the blood or lymphatic vessels. Thus, an increase in the number of CTCs was noted after surgery for stomach cancer, breast cancer, and lung cancer[109-111].

Surgical stress may promote the release of neuroendocrine mediators such as catecholamines and prostaglandins, leading to an increase in immunosuppressive cytokines such as IL-4, IL-10, TGF-β, and vascular endothelial growth factor, as well as proinflammatory cytokines such as IL-6 and IL-8. In turn, increased expression of cytokines and angiogenic and growth factors, as well as suppression of cell-mediated immunity, contributes to tumor invasion and metastasis and the creation of PMNs[112]. In an experiment, perioperative use of a beta-blocker (propranolol) and a COX-2 inhibitor (etodolac) to reduce the levels of catecholamines and prostaglandins reduced liver metastasis of CT26 colon cancer in BALB/c mice[113]. The use of the nonsteroidal anti-inflammatory drug ketorolac in patients with breast cancer during the postoperative period was associated with increased disease-free survival in the first 5 years after surgery and a fivefold reduction in the risk of early relapse[114].

Immunosuppression

One of the mechanisms for activating dormant tumor cells may be immunosuppression. In an experiment, injection of CRC cells into the liver of rats did not lead to the formation of metastases within 60 d. However, injection of cyclosporine A into animals caused liver metastases in 100% of animals and lymph node metastases in 40% of rats within two weeks[115]. Radiation therapy may also promote the awakening of dormant tumor cells and their proliferation and spread through destruction of tumor vessels, induction of EMT and immunosuppression[81].

We also note several promising pathogenetic approaches to therapy aimed at preventing and treating dormant metastases. Considering that increased coagulability and the formation of microthrombi in vessels are important factors necessary for the establishment of metastases, the use of anticoagulants and antiplatelet agents in cancer patients not only prevents thromboembolic complications but also reduces the risk of metastatic organ damage. It is possible that the improved long-term outcome of CRC treatment with aspirin is due to its ability to reduce the risk of blood clots[116,117]. However, in older patients with locally advanced and disseminated CRC, aspirin may worsen long-term treatment outcomes[118]. A Scandinavian multicenter, double-blind, randomized, placebo-controlled trial is currently underway to determine whether adjuvant treatment with low-dose aspirin can improve disease-free survival in patients with liver metastases from CRC[119].

Currently, the first research results have suggested the possibility of “eradicating” dormant tumor metastases. For example, blocking integrin-mediated adhesions between DTCs and the perivascular network increases the sensitivity of dormant DTCs to chemotherapy[10]. Eradication of DTCs may be facilitated by targeting metabolic and adaptive stress signaling pathways such as NRF2[120]. In particular, the possibility of using serine synthesis inhibitors (including PHGDH inhibitors and antifolates) and limiting the supply of exogenous serine in the treatment of CRC has been actively discussed[121,122]. However, the data obtained by Rossi et al[106] are alarming, as they showed that breast cancer with regional and distant metastases was characterized by intratumoral heterogeneity with low PHGDH.

Another interesting and promising direction is the development of drugs aimed at maintaining the dormant DTC phenotype. For example, epigenetic drugs such as azacitidine and retinoic acid, as well as small molecule NR2F1 agonists, can induce DTC dormancy, inhibiting the formation of metastases[123,124].

CONCLUSION

Thus, research has indicated that dormant metastases develop in various malignant neoplasms, including CRCs. Considering the data indicating that: (1) In CRC, dormant metastases develop at the earliest stages of malignant tumor development[8,42]; (2) clusters of tumor cells have a greater ability to form metastases than single tumor cells[94-96]; and (3) the heterogeneity of the primary tumor is associated with a high risk of metastasis[95,102-104], I propose to discuss the possible steps of CRC progression, taking into account the formation of dormant tumor cells.

The long period of remission in patients with early-stage CRC can be assumed to occur because (1) at this stage, the primary tumor has a relatively monoclonal structure; (2) CTCs are represented predominantly by single cells that enter vessels as a result of EMT; and (3) they possess the properties of tumor stem cells. Some CTCs are destroyed by the immune system. However, individual tumor cells can colonize existing dormant niches designed to preserve their own stem cells. In these niches, an appropriate microenvironment can help maintain the dormant state of DTCs for a long time, even in the absence of systemic therapy. The main goal of therapy at this stage is to prevent factors that can activate dormant tumor cells. It is possible that a healthy lifestyle, physical activity, vitamins and antioxidants, and a healthy intestinal microbiota will contribute to this.

As the tumor grows and becomes heterogeneous, both the individual tumor cells and clusters of tumor cells will enter the vascular bed, for example, due to the cavitary type of angiogenesis described above. In addition, complexes of tumor cells can be formed due to the entry of individual CTCs into the corresponding niches, which already contain dormant tumor cells that had entered these niches earlier[44]. At this stage of the disease, long-term remission is possible only with effective systemic drug therapy. This is evidenced by the fact that the prognosis of patients with a complete therapeutic response (RCB0) is significantly better than that of patients with no effect or low effectiveness of treatment (RCBIII-IV)[47,48]. However, effective therapy can likely promote the selective selection of cells with stem characteristics and transfer them to a dormant state. Thus, at this stage of CRC, the main goal of therapy may be to achieve a complete therapeutic response (RCB0) with neoadjuvant drug therapy or to eliminate the majority of cells with an active phenotype with adjuvant therapy. Long-term remission in some patients with locally advanced and even disseminated stage CRC indicates the fundamental possibility of this approach. However, the proposed steps of CRC progression cannot explain why, in some cases, tumors rapidly metastasize with minimal and sometimes undetectable sizes of primary tumors.

We believe that further study of the characteristics of dormant tumor cells and premetastatic niches, the factors influencing the acquisition of the dormant phenotype of CTCs and DTCs, and the activation of dormant metastases will lead to the identification of new promising treatment options and strategies aimed at maintaining controlled remission in patients with CRC.

However, a number of problems are associated with the study of dormant tumor cells, including (1) the lack of adequate experimental models, which makes it difficult to develop new approaches for cancer therapy; (2) the difficulty of clinically assessing the presence of dormant tumor cells, especially since the appearance of clinical metastases can take years; and (3) the lack of uniform approaches for treating dormant metastases[26]. Solving these pressing problems can significantly advance our understanding of cancer biology and the development of effective treatments for malignancies.