探索卵巢癌、胃癌及胰臟癌於腹膜轉移之轉移瓶頸

Abstract

腹膜轉移所面臨的轉移瓶頸在不同癌症類型間仍缺乏系統性的比較，而癌症演化學為此議題提供了更加透徹的洞見。目前仍不清楚，究竟是原發腫瘤中的所有細胞皆具有轉移潛力，還是僅有特定子群的細胞能夠成功產生轉移。在本研究中，分析了三種癌症中腹膜轉移的演化軌跡：透明細胞卵巢癌（Clear Cell Ovarian Cancer）、胃腺癌（Gastric Adenocarcinoma）和胰臟管腺癌（Pancreatic Ductal Adenocarcinoma）。為了量化轉移瓶頸，我開發了一種基於演化樹結構的新指標（Quintet Homogeneity Index，QHI）。結果顯示三種癌症在轉移瓶頸上存在差異：卵巢癌展現出最寬鬆的轉移瓶頸，胰臟癌則最為狹窄，胃癌介於兩者之間。胰臟癌與胃癌這些轉移瓶頸較窄的癌種的腹膜轉移通常會距離（Genetic Distance）原發腫瘤較遠。以區域性淋巴結轉移作為參考時，我們觀察到胰臟癌的腹膜轉移往往來自與淋巴結轉移不同的祖先克隆，顯示其轉移來源更具獨特性；相對地，卵巢癌的腹膜轉移則未出現明顯獨立的克隆，暗示其轉移過程較為寬鬆。在克隆架構分析中，我們發現負責腹膜轉移的祖先克隆，在卵巢癌的原發腫瘤中即已佔有較高比例的癌細胞，進一步呼應其較為寬鬆的轉移瓶頸。了解腫瘤異質性可以幫助我們了解癌症在轉移的各個階段所遇到的瓶頸效應，從離開原位腫瘤到腹膜中開始增生，異質性是透過成對基因距離（pairwise genetic distance）進行評估，我們發現三種癌症的原發腫瘤在異質性上存在顯著差異，且這些差異似乎會影響其腹膜轉移的異質性。然而，三種癌症的腹膜轉移在不同病灶之間的異質性水平相對接近。值得注意的是，復發性腹膜轉移表現出更高的異質性，可能與轉移後細胞分裂次數增加有關。本研究揭示了腹膜轉移過程中癌症類型特異的轉移瓶頸差異，並提出原發腫瘤異質性如何塑造轉移多樣性的初步證據。這些發現為後續探討腹膜轉移演化動態提供了基礎。

A systematic comparison of the metastatic bottlenecks faced by peritoneal metastasis across different cancer types is still lacking, and cancer evolutionary biology offers deeper insights into this issue. Whether all cells in a primary tumor possess metastatic potential, or if only a restricted subset can give rise to metastases. In this study, I analyzed the evolutionary trajectories of peritoneal metastases in three cancers: clear cell ovarian carcinoma, gastric adenocarcinoma, and pancreatic ductal adenocarcinoma. To quantify metastatic bottleneck, I developed the Quintet Homogeneity Index (QHI), a novel metric based on tree topology. The results reveal differences: ovarian cancer exhibits the broadest metastatic bottleneck, pancreatic cancer the narrowest, with gastric cancer falling in between. These findings are further supported by genetic distance, where peritoneal metastases are distant from primary tumor in pancreatic and gastric cancers. Using regional lymph node metastases as a reference, pancreatic cancer peritoneal metastases frequently arise from genetically distinct clones, whereas ovarian cancer peritoneal metastases often share a common origin with lymph node metastases, suggesting a more permissive metastatic route. Clonal architecture analysis showed that the ancestral clones of peritoneal metastases tend to have higher cancer cell fractions in ovarian cancer, consistent with a broad bottleneck. Bottleneck at each step of metastasis, from detachment from the primary tumor to proliferation in the peritoneal cavity, can be captured through tumor heterogeneity. Tumor heterogeneity was assessed via pairwise genetic distances. Primary tumor heterogeneity significantly differed across the three cancer types and appeared to influence the heterogeneity of their peritoneal metastases. However, peritoneal metastases themselves exhibited similar levels of inter-lesion heterogeneity across cancer types. Notably, metachronous peritoneal metastases showed elevated genetic divergence, likely driven by more post-seeding cell divisions. Treatment effects do not alter the pattern of inter-lesion heterogeneity across cancer types. Overall, this study highlights cancer-type-specific differences in the metastatic bottleneck of peritoneal dissemination and presents preliminary evidence for how primary tumor heterogeneity shapes metastatic diversity. These findings provide a foundation for future investigations into the evolutionary dynamics of peritoneal metastasis.