Despite recent advances in the treatment of advanced or metastatic non-small cell lung cancer, distant metastasis remains a major cause of cancer-related death and a barrier to long-term survival. Distant metastasis is an ominous feature and final result of the development of various solid tumors, including NSCLC. The colonization of tumor cells from tumor origin to spreading distant organs is a multi-step but inefficient process. Although we have a good understanding of the cellular and molecular mechanisms involved in distant metastasis, the genetic differentiation and phylogenetic relationships of different metastasis to primary tumors are still largely uncertain.

New evidence suggests that tumor development can be considered an evolving Darwinian model. Metastasis can also be seen as an evolutionary process in which tumor cells spread from primary foci to metastasis through microscopic and/or macroscopic evolution. Unlike the view that classical metastasis only occurs later in the late stage of tumor formation, recent studies have shown that this process may occur early in the formation of tumors. To systematically elaborate this process, two general progress models, namely the linear progress model and the parallel progress model, are proposed.

These two patterns are distinguished from two dimensions: (1) the relative time for metastasis in primary foci; (2) the expected genetic differentiation characterized by comparing the sum of mutations between primary foci and matched metastasis. Although recent comparative studies of primary tumors and metastasis may have linear and parallel metastasis progression at different types of cancer, it is not clear whether metastasis in different organ sites of cancer follows the same or different progression patterns. In addition, heterogeneity within and between metastasis in the same metastasis organ affected by cancer is not fully elucidated. A better understanding of these issues will not only provide new insights into the biology of the metastasis process, but will also potentially reveal differences in treatment strategies for both primary and metastasis.

In order to fully explore the genetic differences and kinship between the two most common metastasis methods of lung adenocarcinoma - liver metastasis (lim) and brain metastasis (brm), we performed whole exon sequencing (wes) of 10 primary tumors (lim-p) and liver metastasis (lim-m), and 11 primary tumors (brm-p) and brain metastasis (brm-m). In order to ensure the natural evolutionary state of the tumor, it will be consistent before any system treatment.

Conditioned tissue and peripheral blood samples were matched. In addition, we collected 16 surgical samples and paired peripheral blood from 5 patients with brms only and sequenced them. Our goals included: (1) determining whether LIM and Brm of lung adenocarcinoma have different mutational landscapes in terms of single nucleotide variants (snv)v of their primary tumors; (2) comparing the evolutionary patterns of LIM and Brm; (3) characterizing the intra-metastatic and inter-metastatic heterogeneity of Brm.

The most typical mutation driver genes included Tp53 (15/23) in 28 (66.7%) samples and EGFR mutations (6/23) in 12 (28.6%) samples (Figure 1a). It is worth noting that Tp53 and EGFR mutations are highly consistent between matched primary and metastatic tumors.

To study intertumor heterogeneity, identified gene changes were classified as shared (presented in primary and metastatic tumors) or private (presented only in primary and metastatic tumors) mutations.23 We observed a significantly higher intertumor homogeneity in the Lim cohort, with the median identified as shared gene mutations at 66.3% (range 6.1%-97.1%). This is in sharp contrast to the median shared gene mutations found in the brm cohort (range 0.0%-30.5%).

TMB is similar between lim-p and lim-m (p>0.999).

We then used Pearson correlation analysis to calculate mutation similarity to describe the consistency and difference in mutation characteristics between paired primary and metastatic tumors. As shown in Figures 1e and 1f, we observed that mutation consistency between lim-p and lim-m in the Lim group, while mutation consistency between brm-p and brm-m in the Brm group was limited. Statistical analysis showed that the Pearson correlation coefficient of mutation similarity in the Lim group was significantly higher than that of the Brm group (p=0.019).

The results show that the lim-p and liv patterns are highly similar, and the pearlson correlation coefficients of 9 of 10 cases (90.0%) were greater than 0.8. However, the paired brm-p and brvs patterns are highly inconsistent, and only 4 of 11 cases (36.4%) had pearlson correlation coefficients of more than 0.8. Statistical analysis shows that the v similarity is significantly higher than that of the brm group (p=0.035).

In total, 42 pathways were significantly enriched in these samples, and compared with Lim-P, Lim-M had a uniquely significantly rich metastatically related pathway, such as notch signaling and ECM-receptor interactions. Notch signaling is also abundant in Brm-M, but not rich in Brm-P. And, Brm-M, there are many unique pathways, such as drug metabolism, several immune-related pathways, apoptosis and mismatch repair. To understand whether the pathway level similarity in Brm-P and Brm-M is also lower than in Lim-P and Lim-M, we performed pathway characteristics analysis and calculated pathway level similarity between sample groups (see Materials and Methods). In fact, Brm-M had significantly different pathway enrichment compared with primary tumors that matched them, while Lim-M had more similar pathways. (p=0.0011).

2. Phylogenetic relationship between paired primary and metastatic tumors

To study the evolutionary process of Lim and Brm (e.g., linear and parallel development model 4), we investigated the phylogenetic relationship between primary tumors and their matching metastatic tumors. We inferred the sequence of genomic alterations and then reconstructed the phylogenetic tree in each case using two independent algorithms. We list three representative cases for each group in Figure 3a.

3. Intra-transfer and intertransfer homogeneity of paired brm

Since Brm showed significantly lower mutation and copy number variation similarity between paired primary foci and metastasis than that of the Lim group, we then studied the intrametastatic and intermetastatic heterogeneity of Brm. We used the method described previously to study the heterogeneity of Brm within (multiple regions of a single brm) and between metastasis (multiple brm from the same patient) in 24 cases, and collected brm specimens from 5 other patients with lung adenocarcinoma. As shown in Figure 4, a single brm lesion was divided into three to four parts from patients brm91, brm93 and brm94 for intrametastatic analysis. To perform intermetastatic heterogeneity analysis, 3 and 2 brm lesions were collected from patients brm95 and brm96, respectively. The most typical mutation driver genes in these samples were egfr and tp53. Similarly, egfr mutations were completely consistent in the samples intrametastatic and intermetastatic samples.

The mutation patterns and characteristics of all samples in the same patient were almost consistent (Figure 4a). In addition, we reconstructed the phylogenetic tree for each case. As shown in Figure 4b, different metastasis samples for all cases can be traced back to a single-line relationship, which suggests that the common origin of multiple brms is dominant.

?discuss?

Despite recent progress in the treatment of various cancers, scientific findings have not been effectively translated into preventive and therapeutic metastasis, which remains the main challenge for long-term survival of treatment failure and frustrating evidence. There is a growing body of evidence that metastasis is partly the product of primary tumor evolution. Clones with strong metastasis may occur early and late in primary tumor progression. A better understanding of the metastasis process and its progression patterns is considered important clinically significant; therefore, these findings may ultimately lead to new strategies for preventing and controlling distant metastasis. In addition, elucidating the unique mutation patterns and evolutionary patterns of different metastasis sites may help to develop precise strategies for this unmet need in clinical practice.

To achieve this goal, we performed wes on 21 lung adenocarcinomas with Lim or Brm to study their mutation patterns and evolutionary patterns. In addition, we conducted multi-regional Wes examinations on another five single-onset Brms to explore heterogeneity within or between metastasis. Our study found that Brm-M exhibited a more different mutation pattern than Lim-M compared to primary tumors. The trajectory of lung adenocarcinomas in Brm and Lim was completely different. Brm is more likely to have a parallel evolution model, while Lim is more likely to have a linear model. Finally, we found that inter-metastasis and within metastasis are common in Brm-M.

We first studied the mutations of Lim and Brm at the SNV level. Although the primary tumors did not differ much in mutation patterns and mutation characteristics from the matched Lim-M or Brm-M, we did notice that Brm-M had a higher proportion of specific cell mutations, which suggests that Brm-M was genetically different from its primary tumor compared with Lim-M. Due to the limitations of SNV, the pattern of V. However, the high consistency of Lim-M and primary tumors was noted, and no such finding was observed in the Brm group. In addition, significantly higher Brv levels than Lim-V are an important indicator of chromosomal instability (cin), and previous literature

It shows that cin is associated with metastasis and the important role of cin in the formation and development of brm deserves further study. Recently, Shih et al. reported that overexpression of myc, yap1 or mmp13 may increase the incidence of brm, suggesting that genome sequencing of a sufficient number of metastatic tumors can find previously unknown drivers of metastasis. In summary, these findings make it reasonable to think that clones that initiate metastasis to different organs may follow different evolutionary paths. This does remind you of Lambert et al. Statement of metastasis ability is not an accidental result of primary tumor progression, but rather a trait selected during primary tumor evolution, as well as the classic “seed and soil” hypothesis.

Early studies of different cancers, including breast and colorectal cancer, showed a high degree of consistency between primary and metastatic tumors, indicating a linear progression model. These results reached a view that the use of sequencing data for primary tumors was used to guide the treatment of metastatic lesions. However, a recent study challenged this paradigm, which found early spread and parallel progression of clones with metastatic ability in primary prostate and metastatic cancer. Similarly, Zhao et al. reported on primary and metastatic tumors (including 13 lung cancers) and found that 11 were following parallel progress. Similarly, Zhao et al.9 reported on 40 pairs of primary swelling.

Tumors and metastatic tumors (including 13 lung cancers), and 11 cases were found to follow parallel progression. In addition, Hosseini et al. found that more than 80% of metastases came from early-spread breast cancer cells, and those early-spread cancer cells were more metastatic than those later-spread cancer cells. Therefore, these results show that parallel and linear progression patterns coexist during cancer evolution. In this study, we found the advantage of linear progression models in Lim, while the advantage of parallel progression models in Brm proposed the need to combine metastases, especially the genetic information of Brm, to guide these patients better tailored treatment.

Finally, we conducted a preliminary analysis of the intrametastatic and intermetastatic heterogeneity of Brm. Despite the small sample size, we observed homogeneity intrametastatic and intermetastatic, as previously reported. In different Brm lesions in the same patient, driver gene changes, mutation patterns and mutation characteristics were found to be almost consistent. Meanwhile, phylogenetic analysis showed that all tumor specimens, whether from the same or different metastatic sites, could be traced back to a single-line relationship, which suggests that multiple Brms have common origins. In addition, we also observed different divergence time points in single metastasis and different Brm lesions. Therefore, we speculated that Brm might indeed have slight microevolutionary transitions, and its biological function is not yet clear. In summary, our study further emphasizes the complexity and uniqueness of Brm evolution.

Taken together, our study reveals that the same primary tumor metastasis to different organ sites may have different mutational landscapes and evolutionary trajectories. These findings will provide new insights into the biology of the metastasis process and provide useful information for the development of new methods to prevent and target distant metastasis.

The findings of this study may have several clinical implications. First, we observed homogeneity of functional mutations in driver genes such as TP53 and EGFR in the Lim and Brm cohorts. This is consistent with recent analysis of genetic heterogeneity in untreated cancers, which also found that 100% of driver mutations in metastatic tumors from the same primary tumor are homogeneous, suggesting a high degree of homogeneity of functional driver gene mutations between paired primary and metastatic lesions. Second, for Lim patients, the mutation status of Lim-M is highly similar to that of their matching primary tumors, suggesting the substitutability of primary tumors to Lim-M when delineating mutation characteristics. However, since Brm-M is more likely to have a parallel evolutionary pattern, we should

It is realized that in a given clinical environment, operable mutations determined by a biopsy of the primary tumor may not represent mutations of brm-m. Third, spatially isolated brain metastases are genetically homogeneous, with significantly higher proportions of branch mutations in brm-m, indicating additional evolution of the brain metastases lineage. Targeted changes in brain metastases represent an important opportunity for new targeted therapeutic strategies to overcome therapeutic resistance and will affect overall survival. Fourth, there are different mutation patterns and evolutionary patterns between pulmonary adenocarcinoma LiM and brm, which means that precise strategies are needed to manage these two different metastases, and also raise the question, namely, to study the evolutionary patterns of other common metastases (such as bone and adrenal metastases).
Chapter completed!