TDLNs, where anti-tumor immune activity is primarily induced, play pivotal roles to effectively generate anti-tumor T cell responses [
18,
19,
35]. TDLNs are usually excised during surgery along with a primary tumor(s) to remove possible residual tumor cells and are examined for diagnosis with/without metastatic tumor cells in these lymph nodes [
36]. It is now well-known that extensive dissection of LNs provided little clinical benefit on the long-term survival for MSI-H/dMMR CRC patients [
37]. It is also suspected that the preservation of lymph nodes is likely to be beneficial for host immune responses against cancer cells [
38]. In this study, we explored the possibility of the use of TDLNs for the ACT and have shown the presence of tumor-reactive T cells in non-metastatic lymph nodes as well as a possibility for expansion of tumor-reactive cytotoxic T cells in vitro by co-culture with autologous tumor cells. As expected, tumor-reactive T cells in lymph nodes in Lynch syndrome (C207) were already activated and proliferated in vivo, lymphocytes needed co-culture with autologous tumor-cells in other samples. The advantage of ACT using TILs is obtaining a large number of T cells that may have anti-tumor activity. The establishment of TILs for ACT usually needs two phases [
39]. In the first phase, lymphocytes were separated from tumor tissues and were cultured in vitro, and then were activated in the medium containing a high concentration of IL-2 to further increase tumor-reactive T cell clones. In the second phase, lymphocytes are further expanded up to 1000 folds (often up to 10
11 cells) in rapid expansion procedure condition [
7,
40]. However, this approach did not work well for advanced-stage CRC patients [
5,
41], but it has been investigated by the combination with immune checkpoint inhibitors [
42]. Exhausted phenotype of CD8
+ T cells were thought to be related to impairment of cytotoxicity [
34]. In fact, TILs showed more exhausted phenotype than lymphocytes after co-culture with autologous tumor cells. On the other hand, it was shown that a smaller number of lymphocytes (median 153 × 10
6 cells) might work effectively for stage IV colorectal cancer patients when lymphocytes were isolated from sentinel lymph nodes [
21]. In any case, the number of tumor-reactive T cells should certainly be one of the key factors for the success of adoptive cell therapy. In our experiences, it is not so difficult to obtain 1 × 10
7 or more lymphocytes from one non-metastatic lymph node, which can be expanded to several times by the 2-week culture with autologous tumor cells. Furthermore, if we combine the rapid expansion procedure in our culture condition, it would be possible to obtain a much larger number of tumor-reactive T cells.
In many cases, cell therapy has been provided to advanced-stage patients who had metastatic, refractory or relapse tumors. In this study, we established autologous tumor-cell lines from fresh tumor tissues. It would be difficult to prepare autologous tumor cell lines in advance for all patients in clinical study. We will get tumor-cell lines more reliably even from cryopreserved tumor tissues if we use organoid culture system [
43]. Moreover, lymphocytes of TDLNs can be kept for a long time after single cell suspension and these cells can be activated, expanded and good cell sources for cell therapy whenever patients experience relapse.
Our data have indicated that levels of anti-tumor-reactivity of lymphocytes after the co-culture with tumor cells might be predicted by the numbers and frequencies of expanded lymphocytes that shared clonotypes with TILs in primary sites. In this study, we were unable to examine the type of antigens that were recognized by lymphocytes proliferated by co-culture with autologous tumor cell lines. We could not also confirm the specificity to tumor cells, but since lymphocytes recognizing self-antigens (except some cancer-testis antigens) are usually eliminated, we assume most or at least some of expanded lymphocytes recognize tumor-specific antigens. Although lymphocytes cultured from lymph nodes would be safe as TILs [
24], it is necessary to check the cross-reactivity with normal cells when applying our approach in the clinic to make sure the safety of the patients. Furthermore, despite we would not be able define the types of lymph node that might have a higher potential as good cell sources for the ACT before the co-culture with autologous tumor cells because we found no meaningful difference in the cell population levels, cell phenotypes or RNA expression profiles. However, we might be able to do this by characterization of T cell populations and phenotypes after the co-culture with tumor cells [
42]. In addition, T cell persistency and proliferation capability in vivo after the cell transfusion seemed to be crucial for the clinical effectiveness of ACT; a T cell stem-like phenotype is possibly related to the in vivo persistency [
44,
45].
Notably, we could enrich tumor-reactive T cells from non-nearest lymph nodes of multiple patients; these lymph nodes did not belong to the nearest regional lymph node group of the tumor. Regional lymph nodes for CRC are categorized as D1, D2 and D3 by anatomical locations. D1 is the nearest and D3 is relatively distant from the primary site. LN#8 and LN#9 of C215 (regional lymph nodes of sigmoid colon tumor of this patient) were located anatomically far from the primary rectal tumor and were not regional lymph nodes of this rectal cancer, but lymphocytes from these lymph nodes showed the immune reactivity against rectal tumor cells. Since the flow of lymphatic vessels is complicated, it is not conclusive that non-metastatic lymph nodes located relatively far from the primary site have been activated by TAAs in vivo. We assume that all lymph nodes, which contain tumor-reactive T cells even at a very low level, can be used as cell sources for ACT therapy when lymphocytes are cultured with autologous tumor cells.
In summary, our findings imply that non-metastatic lymph nodes might be applicable as cell sources for adoptive cell therapy. These data highlighted the utility of non-metastatic lymph nodes for a new therapeutic option for CRC patients with poor prognosis.