探索肝癌的創新療法︰從分子標靶至代謝標靶治療

Abstract

肝細胞癌（簡稱：肝癌）是舉世聞名之高死亡率癌症。蕾沙瓦（sorafenib）是一種抑制血管新生及Raf-1 蛋白的標靶治療藥物，也是目前唯一可延長晚期肝癌患者整體存活期的藥物。然而大多數患者在開始服用 sorafenib後3-4 個月內即出現抗藥性而疾病惡化，面臨缺乏有效治療藥物的困境。且臨床上缺乏偵測抗血管新生能力的生物標記，而無從釐清抗血管新生作用與sorafenib療效的關聯。醫藥界已投入大量資源研究肝癌的分子致癌機轉及研發相關分子之標靶藥物，但是大多數藥物之腫瘤反應率均<10%，顯示肝癌形成並非由單一分子變異引發的，因此同時抑制多種致癌標的或開發其他治療策略是重要的研究方向。 本研究旨在：（1）探討以影像學生物標記偵測抗血管新生標靶藥物之血管新生抑制能力及預測療效之應用，以釐清抗血管新生在肝癌治療的重要性；（2）探討代謝標靶療法在晚期肝癌治療的應用潛力。 標靶藥物之影像學生物標記研究： 動態對比增強核磁共振影像(DCE-MRI)已廣泛應用於抗血管新生標靶藥物之早期臨床試驗以偵測其血管新生抑制能力。本研究針對即將參與一個使用sorafenib併用具抗血管新生能力的節拍式化學治療(tegafur/uracil)的臨床試驗的晚期肝癌患者，取得其知情同意後於治療前及治療14天後進行DCE-MRI 照影。評估治療前後Ｋtrans等指標變化與最佳腫瘤反應、無疾病惡化存活期及整體存活期的關聯。 31位可評估病患的最佳腫瘤反應如下：1位為部分反應(partial response)，18位為疾病穩定(stable disease)，12位為疾病惡化(progressive disease)。達到部分反應及疾病穩定的患者，相對於疾病惡化的患者，其治療前的Ktrans值(中位數：1215.2 × 10-3/min，範圍：582.5∼4555.3 ×10-3/ 分鐘 vs. 中位數：702.0×10-3/min，範圍：375.2∼1938.0 × 10-3/min，P= 0.008)及治療14天後的Ktrans值下降(中位數：–47.1%，範圍：–87% ~ –18% vs. 中位數：9.6%，範圍：–44.8% ~ ＋81%；P <0.001)明顯較大。治療14天後達到『 血管反應(vascular response； Ktrans值下降>=40%)』的患者其中位無疾病惡化存活期(29.1週 vs. 8.7 週，P＝0.033)或中位整體存活期(53.0週 vs. 14.9 週，P＝0.016)均較長。治療後之Ktrans值下降百分比為預測腫瘤反應、無疾病惡化存活期、整體存活期的獨立因子。 本研究結果顯示使用DCE-MRI測量抗血管新生治療引起之早期血管變化有助於預測療效，亦證實抗血管新生作用在肝癌治療的重要性。 代謝標靶治療之研究： 癌細胞偏好使用糖解（glycolysis）作用來因應癌症進展時巨量的合成代謝（anabolism）需求，此種癌細胞特有的代謝表徵（Warburg 效應），已知與癌症進展及抗癌藥物的抗藥性形成有關。雙氯醋酸鹽（dichloroacetate；DCA）為丙酮酸脫氫酶激酶（pyruvate dehydrogenase kinase；PDK）抑制劑，可將糖解轉向氧化磷酸化反應(oxidative phosphorylation)。本研究旨在評估Warburg 效應與肝癌細胞的sorafenib抗藥性的關聯性，以及以DCA改變癌細胞的代謝是否可克服sorafenib抗藥性。 以6株未經sorafenib處理的人類肝癌細胞株及1株(Huh-7)長期暴露sorafenib而篩選出對sorafenib具抗藥性的Huh-7R進行體外測試。所有細胞均在有或無ATP合成酶抑制劑oligomycin處理的情況下測量其乳酸(lactate)產量的變化以推估其利用糖解及氧化磷酸化反應進行合成代謝的比例。以DCA及抑制糖解之己糖激酶( hexokinase 2)之短干擾RNA(siRNA)改變癌細胞代謝。體外測試包括細胞存活率、粒線體膜電位、sub-G1細胞凋亡數、乳酸、活性氧物種(ROS) 、ATP及葡萄糖攝取。以皮下種植對sorafenib有先天抗藥性之人類肝癌細胞株的小鼠，經口餵食vehicle、sorafenib (10毫克/公斤/日)、DCA (100毫克/公斤/日) 或兩者併用，作為評估活體療效的實驗模式。 在所有肝癌細胞株中，越倚賴糖解作用的細胞其對sorafenib的IC50也越高。DCA 可降低乳酸形成，增加ROS及ATP量，顯示DCA確實將糖解轉向氧化磷酸化反應。DCA可回復具sorafenib抗藥性的Hep3B及Huh-7R細胞對sorafenib的敏感度，提升sub-G1細胞凋亡數 （sorafenib 併用DCA vs. sorafenib： Hep3B：65.4±8.4% vs.13±2.9%；Huh-7R：25.3± 5.7% vs. 4.3±1.5%；P值皆<0.0001）。然而 sorafenib併用hexokinase 2之短干擾RNA，相對於sorafenib併用對照之短干擾RNA，只在Huh-7R微幅增加Sub-G1細胞凋亡數 (10.4±1.1% vs. 6.3±0.7%，P=0.0123)，但在Hep3B則無增敏的現象。在動物實驗中， sorafenib 合併 DCA比單用sorafenib明顯抑制腫瘤生長及引發細胞凋亡（治療3周後腫瘤體積相較於控制組：-87% vs. -36%，P值<0.001；治療1周後細胞凋亡量：2.4% v.s 1.2%，P=0.039）。 本研究顯示肝癌細胞偏好使用糖解進行生成代謝的代謝表徵，可以做為預測 sorafenib療效的生物標記以及新的治療標的。 總結及未來展望： 抗血管新生為肝癌的重要治療策略。使用DCE-MRI測量以 sorafenib為基礎之抗血管新生治療引起之早期血管變化有助於預測治療的療效，但DCE-MRI指標對於其他抗血管新生標靶藥物的預測效益仍待進一步探討。而肝癌細胞偏好使用糖解作用進行生成代謝的代謝表徵，可以發展為預測 sorafenib療效的生物標記，其預測效益是否可普及其他藥物則待進一步研究。未來可研發新型的PDK抑制劑做為代謝標靶治療，並進一步釐清活化氧化磷酸化反應與抑制腫瘤生長及增進sorafenib 療效的關係。

Hepatocellular carcinoma (HCC) is associated with a high mortality worldwide as well as in Taiwan. Sorafenib, a molecular targeted therapy inhibiting angiogenesis and Raf-1, is the only drug approved for the treatment of advanced HCC because of its survival benefit. However, most patients develop tumor progression due to sorafenib resistance within 3-4 months after sorafenib use. After sorafenib failure, there are no standard therapies. Moreover, the role of anti-angiogenesis in efficacy of sorafenib remains unclear because of the lack of validated biomarkers for detecting anti-angiogenic activity in HCC patients receiving sorafenib. Many efforts have been spent in investigating molecular mechanisms of hepatocarcinogenesis and developing corresponding molecular targeted agents, but the results strongly suggest that multiple instead of single molecular alterations drive hepatocarcinogenesis. The current study aimed to: (1) explore the role of imaging biomarkers for detection of anti-angiogenic activity of anti-angiogenic molecular targeted therapy and outcome prediction; and (2) explore the potential of metabolism-targeted therapy for advanced HCC. Research in imaging biomarkers of anti-angiogenic molecular targeted therapy: Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) has been widely used in the early-phase trials of anti-angiogenic therapy to detect its biological activity. This study incorporated DCE-MRI examinations before and after 14-day treatment in a phase II trial testing the efficacy of sorafenib plus anti-angiogenic metronomic chemotherapy (tegafur/uracil) if the subjects consented to exploratory DCE-MRI examinations. This study aimed to evaluate the early vascular changes measured by DCE-MRI after 14-day anti-angiogenic therapy and their correlations with the best tumor response, progression-free survival (PFS) and overall survival (OS). Thirty-one patients were evaluable. There were one partial response (PR), 18 stable disease (SD), and 12 progressive disease (PD). Baseline Ktrans was higher in patients with PR or SD (median: 1215.2 × 10-3/min, range: 582.5–4555.3 × 10-3/ min) than patients with PD (median: 702.0×10-3/min, range: 375.2–1938.0 ×10-3/min, P= 0.008). After 14 days of treatment, the median Ktrans change was -47.1% (range:-87.0 to -18.0%) in patients with PR or SD, and -9.6% (range: -44.8% to +81%) in those with PD (P<0.001). A vascular response, defined by ≥ 40% decrease in Ktrans after 14 days of study treatment, correlated with longer PFS (median: 29.1 vs. 8.7 weeks, P= 0.033) and OS (median: 53.0 vs. 14.9 weeks, P=0.016). Percentage of Ktrans change after treatment is an independent predictor of tumor response, PFS and OS. The results suggest that early vascular changes measured by DCE-MRI help predict the efficacy of anti-angiogenic therapy. In addition, the results confirm anti-angiogenesis plays an important role in HCC treatment. Research in Metabolism-targeted Therapy: Cancer cells have a higher propensity for using glycolysis to meet high anabolic demands during cancer progression. This altered metabolism of cancer cells, well known as the “Warburg effect”, has been correlated to anti-cancer drug resistance. Dichloroacetate (DCA), a pyruvate dehydrogenase kinase (PDK) inhibitor, enables to switch glycolysis toward oxidative phosphorylation (OXPHOS). This study aimed to evaluate the correlation between bioenergetic propensity and sorafenib resistance and the potential of DCA to overcome the sorafenib resistance. Six sorafenib-naive HCC cell lines and one sorafenib-resistant HCC cell line (Huh-7R; derived from sorafenib-sensitive Huh-7) were used. The bioenergetic propensity was calculated by measurement of lactate in the presence or absence of ATP synthase inhibitor oligomycin. DCA and siRNA of hexokinase 2 (HK2) were used to target the Warburg effect. In vitro assays included cell viability, mitochondrial membrane potential, sub-G1 fraction, and quantification of lactate, reactive oxygen species (ROS), ATP and glucose uptake. A xenograft mouse model subcutaneously implanted with inherently sorafenib-resistant HCC cells was used for in vivo efficacy. Vehicles, sorafenib (10 mg/kg/day), DCA (100 mg/kg/day) or combination were administered via gavage for 3 weeks. The bioenergetic propensity for using glycolysis correlated with decreased sorafenib sensitivity. DCA reduced lactate production and increased ROS and ATP, indicating metabolism switch toward OXPHOS. DCA sensitized sorafenib-resistant Hep3B and Huh-7R cells to sorafenib-induced apoptosis [Sub-G1 (combination vs. sorafenib): Hep3B, 65.4±8.4% vs.13±2.9%; Huh-7R, 25.3± 5.7% vs. 4.3±1.5%; each P<0.0001], whereas siRNA of HK2 only sensitized Huh-7R cells [Sub-G1 (combination vs. sorafenib plus negative control siRNA): 10.4±1.1% vs. 6.3±0.7%, P=0.0123], but not Hep3B. Sorafenib (10 mg/kg/day) plus DCA (100 mg/kg/day) resulted in superior tumor regression and apoptosis than sorafenib alone in mice (tumor size after 3-week treatment: -87% vs. -36%, P<0.001; apoptotic cells after 1-week treatment: 2.4% v.s 1.2%, P=0.039). The results suggest that bioenergetics propensity of HCC may be a useful predictive biomarker of sorafenib sensitivity and a novel therapeutic target. Conclusions and future perspectives: Anti-angiogenesis is an important therapeutic strategy for advanced HCC. Early vascular changes measured by DCE-MRI following anti-angiogenic therapy can help predict the clinical outcomes. The predictive value of DCE-MRI parameters for other anti-angiogenic combination therapy in HCC patients remains to be determined. The bioenergetic propensity is a potentially useful predictive biomarker for sorafenib sensitivity, but its predictive values for other anti-cancer drugs remain unknown. Future research may focus on developing novel PDK inhibitors and exploring how activation of OXPHOS mediates cancer cell death and ameliorate sorafenib resistance.