Onward Spread from Liver Metastases Is a Major Cause of Multi-Organ Metastasis in a Mouse Model of Metastatic Colon Cancer.

Colorectal cancer metastasizes predominantly to the liver but also to the lungs and the peritoneum. The presence of extra-hepatic metastases limits curative (surgical) treatment options and is associated with very poor survival. The mechanisms governing multi-organ metastasis formation are incompletely understood. Here, we tested the hypothesis that the site of tumor growth influences extra-hepatic metastasis formation. To this end, we implanted murine colon cancer organoids into the primary tumor site (i.e., the caecum) and into the primary metastasis site (i.e., the liver) in immunocompetent mice. The organoid-initiated liver tumors were significantly more efficient in seeding distant metastases compared to tumors of the same origin growing in the caecum (intra-hepatic: 51 vs. 40%, p = 0.001; peritoneal cavity: 51% vs. 33%, p = 0.001; lungs: 30% vs. 7%, p = 0.017). The enhanced metastatic capacity of the liver tumors was associated with the formation of 'hotspots' of vitronectin-positive blood vessels surrounded by macrophages. RNA sequencing analysis of clinical samples showed a high expression of vitronectin in liver metastases, along with signatures reflecting hypoxia, angiogenesis, coagulation, and macrophages. We conclude that 'onward spread' from liver metastases is facilitated by liver-specific microenvironmental signals that cause the formation of macrophage-associated vascular hotspots. The therapeutic targeting of these signals may help to contain the disease within the liver and prevent onward spread.

Lymphatic Invasion of Plakoglobin-Dependent Tumor Cell Clusters Drives Formation of Polyclonal Lung Metastases in Colon Cancer.

BACKGROUND & AIMS: Patients with colon cancer with liver metastases may be cured with surgery, but the presence of additional lung metastases often precludes curative treatment. Little is known about the processes driving lung metastasis. This study aimed to elucidate the mechanisms governing lung vs liver metastasis formation. METHODS: Patient-derived organoid (PDO) cultures were established from colon tumors with distinct patterns of metastasis. Mouse models recapitulating metastatic organotropism were created by implanting PDOs into the cecum wall. Optical barcoding was applied to trace the origin and clonal composition of liver and lung metastases. RNA sequencing and immunohistochemistry were used to identify candidate determinants of metastatic organotropism. Genetic, pharmacologic, in vitro, and in vivo modeling strategies identified essential steps in lung metastasis formation. Validation was performed by analyzing patient-derived tissues. RESULTS: Cecum transplantation of 3 distinct PDOs yielded models with distinct metastatic organotropism: liver only, lung only, and liver and lung. Liver metastases were seeded by single cells derived from select clones. Lung metastases were seeded by polyclonal clusters of tumor cells entering the lymphatic vasculature with very limited clonal selection. Lung-specific metastasis was associated with high expression of desmosome markers, including plakoglobin. Plakoglobin deletion abrogated tumor cell cluster formation, lymphatic invasion, and lung metastasis formation. Pharmacologic inhibition of lymphangiogenesis attenuated lung metastasis formation. Primary human colon, rectum, esophagus, and stomach tumors with lung metastases had a higher N-stage and more plakoglobin-expressing intra-lymphatic tumor cell clusters than those without lung metastases. CONCLUSIONS: Lung and liver metastasis formation are fundamentally distinct processes with different evolutionary bottlenecks, seeding entities, and anatomic routing. Polyclonal lung metastases originate from plakoglobin-dependent tumor cell clusters entering the lymphatic vasculature at the primary tumor site.

Liver lymphatic drainage patterns follow segmental anatomy in a murine model.

The liver's cellular functions are sustained by a hierarchical, segmentally-organized vascular system. Additionally, liver lymphatic vessels are thought to drain to perihepatic lymph nodes. Surprisingly, while recent findings highlight the importance of organ-specific lymphatics, the functional anatomy of liver lymphatics has not been mapped out. In literature, no segmental or preferential lymphatic drainage patterns are known to exist. We employ a novel murine model of liver lymphangiography and in vivo microscopy to delineate the lymphatic drainage patterns of individual liver lobes. Our data from blue dye liver lymphangiography show preferential lymphatic drainage patterns: Right lobe mainly to hepatoduodenal ligament lymph node 1 (LN1); left lobe to hepatoduodenal ligament LN1 + LN2 concurrently; median lobe showed a more variable LN1/LN2 drainage pattern with increased (sometimes exclusive) mediastinal thoracic lymph node involvement, indicating that part of the liver can drain directly to the mediastinum. Upon ferritin lymphangiography, we observed no functional communication between the lobar lymphatics. Altogether, these results show the existence of preferential lymphatic drainage patterns in the murine liver. Moreover, this drainage can occur directly to mediastinal lymph nodes and there is no interlobar lymphatic flow. Collectively, these data provide the first direct evidence that liver lymphatic drainage patterns follow segmental anatomy.

A Giant Brunneroma Causing Gastrointestinal Bleeding and Severe Anemia Requiring Transfusion and Surgery.

Brunner's gland hamartoma, also called hyperplasia, adenoma, and Brunneroma, is an extremely rare benign proliferative lesion of Brunner's glands in the duodenum. While being mostly small and asymptomatic, they can result in gastrointestinal bleeding and obstruction. We report the case of a 54-year-old man presenting with melena and severe anemia requiring blood transfusion. CT scans showed a large mass of 8 cm in diameter, presumably arising in the duodenum. Endoscopic biopsies were not conclusive. As we were unable to determine the nature of the mass preoperatively and due to the severe symptoms, its size, and the uncertain malignant potential, a classic Whipple procedure was performed. The resected specimen showed extensive proliferation of Brunner's glands without signs of malignancy.