【2/2】

【Tomorrow 10-11 o'clock】

There are much less research on the prognostic role of HGP in non-colorectal cancer liver metastasis. These patterns have been described in uveal melanoma, gastric cancer and breast cancer liver metastasis. In these cases, the indications for hepatic resection are less direct, and it is possible that accurate prediction of HGP can save some candidate patients from potentially morbid and useless hepatic resection.

Regarding liver metastasis and uveal melanoma in breast cancer, it appears to be in a replacement mode. This is consistent with the fact that antiangiogenesis treatment failed in clinical trials for breast cancer, which is predictable because most lesions are substituted and judging from the Crclm data, these lesions do not respond to this treatment.

Strangely, in our experience, liver metastasis in pancreatic bile-denocarcinoma also exhibits the patterns described by Crclm, but their impact on prognosis remains to be determined, as this is an extremely rare indication for hepatic resection and any clinical cohort may be very small.

The spread of HGP in studies on liver metastasis in non-colorectal cancer opens up broad possibilities and offers more interest in the prediction of HGP. However, the biological mechanisms involved are not necessarily the same as CRCCLM. For example, the associations between driver mutations should be less consistent, and we expect that the brca (breast cancer) 1/2 and Braf mutations are ultimately associated with HGP in liver metastasis in breast cancer and melanoma, respectively, taking into account the importance of these genes in these tumors. We expect that the biological properties—angiogenesis and vascular co-selecting mechanisms should be similar. However, this has not been confirmed.

Finally, another clinical significance of HGP is the selection of targeted therapies for CRCCLM, which are currently based on epidermal growth factor receptor (EGFR) blockade (cetuximab or panitumab) or vascular endothelial growth factor (VEGF) blockade (bevacizumab, an anti-angiogenic drug). Which one is chosen usually depends on the presence of the RAS mutation, which excludes the efficacy of anti-EGFR antibodies and favors anti-VEGF drugs. Nevertheless, considering that connective tissue hyperplasia HGP is often associated with more angiogenesis, predicting this pattern may contribute to drug selection

The results of the study of the Lazaris and colleagues describe the reduction of microvascular density after bevacizumab combined with connective tissue hyperplasia HGP for Crclm, which is a very promising result, and vascular selection is already a known mechanism for the development of resistance to anti-angiogenic drugs by hepatocellular carcinoma. In a 2017 study from Kather et al., in a small cohort of Crclm, it described a blood-rich area at the junction with surrounding liver tissue, but no correlation with HGP. More studies are clearly needed to address this problem.

Nevertheless, in the field of targeted therapy, it would be very interesting to replicate the results of Wu et al. They described the association between the PIK3CA mutation in primary colorectal cancer and the connective tissue hyperplastic HGP in Crclm. In this case, it is promising to evaluate the presence of PIK3CA mutations in Crclm. Regarding chemotherapy, PIK3CA mutations have been described as mechanisms for the development of resistance to first-line drugs through stem cell survival and proliferation, but recent discoveries in breast cancer have opened up new perspectives and possibilities for PI3K inhibitors. In this case, CRCLM's HGP can also be used as an indication for individualized treatment. Of course, this is purely theoretical.

In the age of nanotechnology, the development of nanocarriers is seen as a promising approach because it will deliver drugs of interest directly in tumors, thus bypassing the side effects of conventional drug registration. Interestingly, HGP may have an impact on nanotherapy, and connective tissue HGP represents a physical barrier to drug delivery, strengthening the importance of understanding them before treatment decisions.

Finally, the most advanced treatments for colorectal cancer at present means microsatellite instability (MSI) screening of patients, which are performed in some institutions in all patients. MSI patients, also known as MSI high, are approximately 10% of all patients with colorectal cancer, can receive immunotherapy. It is well known that these patients have a good prognosis and so far MSI has not been associated with HGP, which would be very interesting.

Characteristics of primary tumors

HGP is considered a phenotypic result of liver tumor-host interactions. However, it seems logical that certain characteristics of primary tumors may drive the biological behavior of CRCCLM and be responsible for HGP. Since many patients suffer from metachronic diseases, correct and detailed research on tumors may provide valuable information.

Several studies have observed that primary colorectal cancer also exhibits different growth patterns. Koelzer and Lugli define invasive tumor growth (the anatomy of tumor tissue to the intestinal wall structure with little or no connective tissue stromal response) and propulsive tumor growth (the clear demarcation of tumor invasion front and host tissue). This finding is consistent with other studies that identify invasive growth patterns and link invasive growth patterns with the worst outcomes and invasive disease (i.e., higher risk of lymph node metastasis).

In primary tumors, just like in Crclm, this invasion pattern is thought to represent tumor-host interactions regulated by molecular alteration, tumor expansion and invasion, and is associated with higher microvascular density in invasive patterns. A study by Rajaganeshan et al. In 2007, he described that propellant growth colorectal cancer preferentially with enveloped liver metastasis, while invasive colorectal cancer is associated with poorer boundary lesions. At that time, there was no consensus on the classification of Crclm, but this finding may suggest that tumor marginal or tissue growth patterns may be similar characteristics of primary and secondary tumors.

In the study of Koelzer and Lugli, they were also able to link this feature to molecular features, such as Braf mutations and mismatch repair defects in invasive patterns and propulsive growth patterns. The correlation between growth patterns and specific molecular changes also provides clear evidence that molecular pathways are activated and controls the growth dynamics of colorectal cancer, with prognostic and therapeutic significance.

Recently, in a study conducted by Wu et al., this possibility was also explored in 2019. In 29 paired primary colorectal cancer and Crclm patients, the presence of dilated growth patterns, low tumor germination and Crohn-like reactions were associated with the proliferative tissue hyperplasia of Crclm. The invasive growth patterns of primary tumors, medium to high tumor germination and lack of Crohn-like reactions were associated with alternative HGP in liver metastasis. Interestingly, in five patients, they were able to perform complete exome sequencing (Wes) and found that Crclm's HGP was associated with mutations in ACP and Pik3CA.

In liver oncology, preoperative imaging and cross-sectional techniques are necessary. Computed tomography (CT) and magnetic resonance (MR) with contrast enhancement can provide information on the location and number of nodules, as well as the vascular relationships of nodules, and facilitate decision-making and surgical planning.

In addition to diagnosis and staging, CT-based imaging also provides relevant prognostic information, i.e., when analyzing the tumor-liver interface. However, in two studies that found clear radiologic boundaries between nodules and non-tumor liver parenchyma, there was no correlation with HGP. Recently, a study showed that pro-connective tissue hyperplasia HGP showed obvious imaging features in enhanced CT, such as a clear margin of portal phase reinforcement between nodules and normal liver parenchyma. In this study, radiomic-based CT image analysis greatly improved the accuracy.

Liver-specific contrast MR imaging is the current gold standard for imaging CRCCLM patients. It provides higher sensitivity for smaller nodules, can assess chemotherapy-related and ischemia-reperfusion liver damage, and can also measure preoperative hepatocyte function. Recently, liver magnetic resonance has accurately predicted the presence or response to chemotherapy, paving the way for the inclusion of imaging characteristics as a non-invasive biomarker.

Given its routine clinical application and advantages in soft tissue analysis, liver-specific contrast enhancement MR has the potential to predict HGP of specific Crclm very accurately and therefore potentially become a reliable non-invasive prognostic biomarker. Further studies are necessary to explore these possibilities. An example of a radiological approach to this issue can be seen in Figure 6.

In the Crclm HGP study, an interesting finding was that most liver nodules showed the same HGP. This finding reinforced that HGP is the phenotype result of tumor-host interaction. With this in mind, HGP is expected to remain unchanged in subsequent hepatic resections. A study by galjart et al. showed that 66 cases of first hepatic resections were compared with the second hepatic resection to observe the overall maintenance of HGP. This finding could be for patients who chose the second hepatic resection, or even for more

Good drug treatment is very helpful. Now, after the multidisciplinary tumor committee makes a treatment decision, patients are selected for additional hepatic resection based on the patient's condition and response to chemotherapy. If the patient is selected correctly, the second and third hepatic resections are considered safe and feasible. Predicting the possibility of HGP in the second hepatic resection will allow better patient choices in this regard, while applying the correct neoadjuvant-antiangiogenic drug in angiogenic HGP                                                                                                                                                                                                

Although HGP itself cannot be evaluated by biopsy, some data can be inferred. In a 2012 study of 24 CRCCLM by Eefsen et al., the expression of urokinase-type plasminogen activator receptor (upar) in connective tissue hyperplasia metastasis tumors was significantly increased. In 2015, the same research team confirmed in a large cohort of 254 CRCCLM, higher upar expression in patients with elevated upar expression in connective tissue hyperplasia HGP and A-upar.c.c.

D3 is low in mixed Crclm expression and does not differ in non-NA patients. This finding can be perfectly extrapolated to biopsy samples and can play an additional role in the study of immune populations because it predicts response to immune checkpoint inhibitors. Compared with alternative HGP, connective tissue proliferation HGP has the characteristics of dense lymphocyte infiltration and high expression of immune-related genes Cd8a, Cd8b, GZMA, GZMB and PRF1, and is a valuable immunomodulatory therapy biomarker.

Detailed analysis of immune infiltration may also allow the microenvironmental characteristics of tumor cells and cancer-related fibroblast release, especially cytokine levels and production, to be fully characterized. The role of macrophages in cancer metastasis is also important. The conversion of the m2 phenotype (pro-tumor) to the m1 phenotype (anti-tumor) supports vascular normalization and stimulates the anti-tumor immune response. Several studies have discussed the role of cytokine and colorectal cancer. As Bazzichetto and his colleagues have reviewed, higher levels of il-8 are associated with colorectal cancer disease progression, liver metastasis

As well as Braf and Pten mutations. Pretzch et al. conducted a very interesting review of the tumor microenvironment and metastasis typing, pointing out that some micrororna (mirna) are associated with specific organ spread in colorectal cancer, namely high expression of mirna-31-5p and low levels of mirna-31-5p and liver metastasis. It would be interesting to see if there is an association between specific mirna and il and HGP, which will make the determination of biomarkers in the serum possible.

Recent studies have addressed the differences between vascularization and HGP. A study by Lazaris et al. showed in 2018 that, compared with the promoproliferative HGP, the microvascular density of replacement HGP at the tumor/nontumor parenchymal interface is higher, and the blood vessels are continuous with sinusoidal vessels, indicating a mechanism of vascular co-selectivity. In the promoproliferative HGP, the blood vessels are continuous with the arterioles, representing angiogenesis. If non-tumor tissue exists, these findings can be replicated in biopsy. An interesting finding in this study is that in the analysis

During tumor microvascular density, the value of connective tissue proliferation HGP is significantly reduced after bevacizumab treatment, which strengthens the role of these drugs in CRCCLM treatment and the necessity of predicting HGP. A recent study conducted by Stremitzer et al.: The relationship between this association in connective tissue proliferation HGP and angiogenic behavior was demonstrated in 118 patients, and evidence of inflammatory immunophenotypes has been shown, which prompts HGP not only for perioperative bevacizumab-based chemotherapy, but also for immuno-targeted therapy.

These findings describe a higher number of expression of angiopoietin 1 associated with the angiocoselective mechanism in hepatocytes adjacent to CRCCLM with alternative HGP even after treatment with chemotherapy and/or bevacizumab. Once again, this finding has the potential to be applied to biopsy. Other markers associated with angiogenesis and hypoxia have been described in colorectal cancer, i.e., the high expression of hypoxia-induced factor (HIF) and carbonic anhydrase ix (CA9).

These markers are also likely to be applied to tissues obtained in biopsy. Liquid biopsy has appeared in the environment of colorectal cancer. Its main role may be to determine the RAS status in patients with IV stages, and the compliance rate is close to 90% when the tissue is unavailable. It also has a very interesting effect on the follow-up assessment of the dynamic changes in KRAS mutation status in patients with wild-type colorectal cancer. The results of Wu et al. are integrated. Specific mutations, such as PIK3CA, can identify characteristics in liquid biopsy and provide information that may be related to Crclm     HGP. This would be a very interesting and less invasive approach, especially for patients undergoing colorectal cancer surgery at another institution where information that can predict Crclm    HGP is not available or difficult to obtain.
Chapter completed!