At present, with the continuous research on anti-tumor immunity, immunotherapy–especially with respect to immune checkpoint inhibitors led by anti-PD-1/PD-L1–has become a widely accepted therapy in clinical practice [1]. At this stage, immunotherapy targeting immune cells, particularly that which focuses on T cells, has emerged as a powerful weapon against tumors, and includes immune checkpoint blockades, adoptive cellular therapies, and cancer vaccines [2]. However, a large number of cancer patients do not benefit from this novel and effective treatment. This is mainly due to the complexity and heterogeneity of the tumor microenvironment (TME) and the diversity of the immunomodulatory network [3]. The intensity of PD-L1 expression on tumor cells and the mutational burden of the tumor have been considered as biomarkers for determining the population that is effective in immunotherapy. In addition, tumor-infiltrating lymphocytes (TILs) have been shown to be highly correlated with efficacy, but they have not been used as a predictive biomarker for patient selection ^[4][5][4,5]. Although the efficacy of immunotherapy combined with other treatments in populations without PD-L1 expression has been demonstrated in previous studies [6], the mechanisms need to be explored further. Some studies have shown that tertiary lymphoid structures (TLSs)–defined by clusters of immune-infiltrating cells–in tumors or the tumor periphery are mostly correlated with better prognosis in patients, whether or not PD-L1 is expressed [7]. A TLS is a spatial structure composed of B cells in the center, surrounded by T cells and a variety of immune cells and immune-related cells, with no envelope covering its surface. Similar to SLOs, TLSs regulate the immune microenvironment by recruiting circulating immune cells and enhancing local immune action. As sites for the generation of circulating immune cells that control tumor progression and better prognosis, independent of the expression of PD-L1, TLSs hold great prospects and potential. In order to better understand what kind of TLSs may benefit patients and the role of TLSs in tumor immunotherapy, in this papentry researchersr we summarize the overall roles of TLSs and their components, as well as clinical trials using TLSs as markers and related research in recent years.

2. The Function of TLSs as Complete Structures

Tertiary lymphoid structures, which are also known as ectopic lymphoid organs, are aggregates of immune cells that appear in tissues where no secondary lymphoid organ (SLO) exists. According to a large number of studies in the past year, TLSs are usually observed at inflammatory sites in response to autoimmune diseases, infectious diseases, organ transplantation, inflammatory disorders, and tumors. There is mounting evidence suggesting their formation is closely related to immune responses mediated by exposure to chronic inflammation. TLSs usually lead to a persistent autoimmune response with a negative impact on subsequent pathological conditions in other diseases ^[8][9][10][8,9,10]. However, TLSs are still considered to be a favorable prognostic and predictive factor in tumors ^{[11][12][13][14][15][16]}[11,12,13,14,15,16]. TLSs are able to provide local anti-tumor immunity, independent of SLOs, which has been demonstrated in the absence of SLOs in mouse models [17]. At present, TLSs can be detected by methods including H&E, immunohistochemistry (IHC), multiplex immunofluorescence (IF) techniques, various chemokines/cytokines of TLS detection, and related gene expression [18]; however, there is no TLS detection and counting criterion that is widely accepted in clinical practice. On H&E-stained slides, TLSs were identified morphologically as distinct ovoid lymphocytic aggregates presenting HEV and/or a germinal center [19]. A TLS has been described in detail as a germinal center consisting of CD20+B cells and plasma cells, surrounded by CD3+T cell zones with DCs and peripheric HEVs, resembling SLOs in formation, structure, and function ^[18][20][18,20]. Therefore, SLOs are the most suitable model by which to better understand how TLSs are formed; SLOs have been proven to have many similarities with TLSs in preclinical mouse models [21]. The genesis of TLSs results from a highly ordered sequence of events involving interactions between hematopoietic and non-lymphoid stromal cells, in which cytokines, chemokines, adhesion molecules, and survival factors including CXCL13, CXCL12, CCL19, CCL21, the tumor necrosis factor superfamily, IL-6, and IL-7 ^[22][23][22,23] play key roles as molecular components [24]. Additionally, one study recently demonstrated that TGF-β-mediated silencing of STAB-1 that induces Tfh cell differentiation and the subsequent formation of intra-tumor TLSs further sheds light on the origin of TLSs in complex TME [25]. TLS density, location, and maturation have been demonstrated to have close associations with favorable clinical outcomes ^{[19][26][27][28][29][30]}[19,26,27,28,29,30], including disease-free survival (DFS) and overall survival (OS). TLS density is correlated with germinal center formation and the expression of genes associated with adaptive immune response, and it is a strongly independent prognostic marker in lung cancer, colorectal cancer, pancreatic cancer, and breast cancer ^{[31][32][33][34]}[31,32,33,34]. There have also been opposing observations that TLS density has no or limited relation with better survival ^[35][36][35,36]. In patients with lung squamous cell carcinoma who received neoadjuvant chemotherapy, the density of TLSs in tumors became similar, losing its predictive effect [27]. The high TLS density in tumors is usually associated with high TIL density as well as with the expression of PD-1/PD-L1, suggesting the potential benefit from immunotherapies ^[37][38][37,38]. However, it was found that a subtype of glioma with high immune infiltrations and TLSs resulted in poor prognosis. This may be because the beneficial effect of TLSs could be reduced in patients with higher immunosuppressive cells such as myeloid-derived suppressor cells (MSDC) [39]. With regard to TLS location, many studies have reported the persistence of TLSs in tumors or peritumors. Peritumoral TLSs often play a favorable role in prognosis in lung cancer [27], pancreatic cancers [40], CRC [41], oral squamous cell carcinoma [42], metastases of melanoma [43], and ovarian cancer [44], but they are also associated with poor prognosis and invasive metastasis in breast cancer [19]. Intra-tumoral TLSs are described as a favorable maker in lung cancer [45] and HCC [46]. With regard to TLS maturation, according to the works of Silina et al., the development of tumor-associated TLSs follows sequential stages of maturation: (1) early TLSs (E-TLSs), T cells, B cells, and CXCL13-expressing perivascular cells gather into clusters without FDC; (2) primary follicle-like TLSs (PFL-TLSs), i.e., TLSs containing FDC without GCs; 3) secondary follicle-like TLSs (SFL-TLSs), i.e., TLSs are analogous to the SLO follicles [27]. E-TLSs lacking germinal centers may favor immune evasion and progression to full-blown HCC in liver precancerous lesions. The presence of mature TLSs is also associated with an improved objective response rate, progression-free survival, and overall survival independently of PD-L1 expression status and CD8+ T-cell density [7]. In general, TLSs vary widely in density, location, maturation, and components, as well as in their proportions in different individuals, and this high heterogeneity needs better classification and further research for application in clinical practice. As mentioned above, although TLS plays a positive role in anti-tumors immunity in the vast majority of reports, there are still some studies that report its negative effects. Peri-cancerous TLS was considered a major contributor to adverse effects on prognosis [19]. One study showed the abundance of Treg cells in intra-tumoral TLS increased significantly with the increase in peri-cancerous TLS regions [47]. This suggested that Treg cells in peri-cancerous TLS may be responsive to suppression of the anti-tumor response and the interplay between peri-cancerous and intra-tumoral immune cells. An additional prognostic disadvantage was immature TLS. In immature TLS, the types and numbers of immune cells vary greatly, such as B cells, which were low and produced immunosuppressive cytokines in immature TLS [48]. In addition, TLS was found to be associated with immune-related adverse events [49].

3. The Role of TLS Components in Tumor-Specific Immune Response

TLSs, as important factors associated with a series of anti-tumor specific immune responses, play multiply significant roles in tumor progression and suppression in TME. Earlier research has suggested that the double roles of each TLS might differ depending on their composition [50]. The cellular components of TLSs affect the function of the anti-tumor immune response in different types of cancer. Therefore, to further understand the prognostic value of TLSs and their dual role in anti-tumor immunity, differences in TLS components and their ratios need to be considered. Next, rwesearchers focus on five major components of TLSs: T cells, B cells, DCs, HEVs, and TLS-associated cells (Figure 1).