Epigenetic Therapy for Thoracic Malignancies

我们发表的关于100多例肺癌、食管癌、胸膜间皮瘤或非胸部恶性肿瘤肺转移患者的研究清楚地表明，地西他滨和罗米地辛单独或联合使用可以调节胸部恶性肿瘤的基因表达，并诱导对这些肿瘤的免疫反应。我们的目标一直是结合表观遗传启动方案与过继免疫治疗胸部恶性肿瘤。目前，不良的生物分布和全身毒性阻碍了对胸部恶性肿瘤进行重编程所必需的表观遗传药物的最佳、慢性管理。例如，DNA去甲基化剂作为口服给药时，生物利用度非常有限，半衰期极短（5min），由于胞苷脱氨酶（CDA）在全身水平很高，导致实体肿瘤吸收不良。为了克服这些局限性，我们与克利夫兰诊所合作，配制了DAC和四氢吡啶（一种有效的、无毒的CDA抑制剂）口服给药。目前正在进行一项口服DAC/THU和派姆单抗治疗不能手术的NSCLC、食管癌或恶性胸膜间皮瘤的I/II期研究。几乎所有患者都表现出系统性表观遗传重编程的证据，并且在一些患者中观察到令人印象深刻的，近乎完全和持久的（1年）反应。为了进一步优化肺恶性肿瘤免疫检查点阻断的表观遗传启动，同时降低潜在的全身毒性，我们最近启动了一项I/II期试验，研究通过吸入技术联合M7824（一种双重免疫检查点抑制剂- tgf - β陷阱）对局部晚期肺转移患者的毒性和潜在疗效。目前在世界其他地方没有开展类似的工作；该试验旨在建立评估一系列雾化表观遗传药物单独或与过继免疫疗法联合治疗局部晚期肺部恶性肿瘤的范例。尽管癌睾丸抗原在多种人类恶性肿瘤中表达，但由于低水平、异质抗原表达、抗原加工/呈递缺陷以及局部和全身免疫抑制，对这些蛋白的免疫反应在胸部肿瘤患者中并不常见。我们发表的细胞系和患者活检研究表明，胸部恶性肿瘤表现出不同的CTA表达模式，以及对上调CTA的表观遗传方案的异质反应。克服这些限制的一个策略是使患者免疫一组cta，这些cta可能通过全身给药染色质重塑剂而上调。为了解决这个问题，我们进行了一项2.5期的首次人体试验，以确定肿瘤裂解物疫苗是否能诱导对CTA的广泛免疫，并确定口服环磷酰胺和塞来昔布（cy/ cell）是否能增强疫苗诱导的免疫反应。简而言之，21例原发性胸部恶性肿瘤或胸外肿瘤转移至胸部，经常规治疗导致NED的患者随机接受H1299肺癌细胞裂解液和Iscomatrix佐剂，通过深皮内注射，每个月x 6 +/-每天cy/ cell。主要终点是第六次接种后1个月对纯化CTA的血清学反应。探索性目标包括分析对碳水化合物抗原的血清学反应性以及评估疫苗治疗前后的外周免疫亚群。所有患者在接种疫苗后72-96小时均表现出局部和全身炎症反应。没有剂量限制治疗相关的毒性。14名患者（67%）完成了所有六种疫苗接种。中位随访时间为64.6个月（58.8-68.4个月），12例患者存活并患有NED；2例患者存活，第二恶性肿瘤得到控制。14名存活患者中有11名（79%）接种了6次疫苗，而7名患者中有3名（42%）死于恶性肿瘤。10名随机接种cy/cel疫苗的患者中有7名接受了全部六种疫苗接种；7例患者中3例（43%）死于疾病。11例随机接受单独疫苗接种的患者中，有7例完成了6次疫苗接种；7例患者无死亡（0%）。8例患者（57%）对NY-ESO-1表现出血清学反应。对GAGE7、XAGE和MAGE-C2观察到其他反应性。对肿瘤相关碳水化合物抗原的反应并不常见。疫苗治疗降低了Tregs （p=0.067*）、Tregs上PD-1表达（p=0.023*）和CD14+单核细胞（p=0.0089*）、经典单核细胞（p=0.0159*）和中间单核细胞（p= 0.0031*）上PD-L1表达的百分比。Cy/cel不影响免疫应答或疫苗诱导的外周免疫亚群的改变。反应的实验室指标似乎与患者生存无关。这项阳性试验的结果最近在网上发表，并支持在胸部恶性肿瘤患者中进一步开展疫苗接种工作。

Our published studies pertaining to more than 100 patients with lung and esophageal cancers and pleural mesotheliomas, or pulmonary metastases from non-thoracic malignancies have clearly demonstrated that Decitabine and romidepsin alone or in combination can modulate gene expression in thoracic malignancies and induce immune responses against these neoplasms. Our goal has always been to couple epigenetic priming regimens with adoptive immunotherapy for thoracic malignancies. Presently, poor biodistribution and systemic toxicities prevent optimal, chronic administration of epigenetic agents necessary to reprogram thoracic malignancies. For example, DNA demethylating agents have very limited bioavailability when administered as oral agents and extremely short half-lives (5min) resulting in poor uptake into solid tumors due to cytidine deaminase (CDA) which is present at high levels throughout the body. To overcome these limitations, we formulated DAC and tetrahydrouridine (a potent, non-toxic inhibitor of CDA) for oral administration in collaboration with the Cleveland Clinic. A phase I/II study of oral DAC/THU and pembrolizumab for patients with inoperable NSCLC, esophageal cancers, or malignant pleural mesotheliomas is currently underway. Virtually all patients have exhibited evidence of systemic epigenetic reprogramming, and impressive, near complete and durable (1 yr) responses have been observed in several patients. To further optimize epigenetic priming of pulmonary malignancies for immune checkpoint blockade while decreasing potential systemic toxicities, we have recently initiated a phase I/II trial to examine the toxicities and potential efficacy of AZA administered via inhalation techniques in combination with M7824, a dual immune checkpoint inhibitor-TGF-beta trap in patients with locally advanced pulmonary metastases. No similar efforts are currently underway elsewhere in the world; this trial is intended to establish the paradigm for evaluation of a series of aerosolized epigenetic agents alone or in combination with adoptive immunotherapy for the treatment of locally advanced pulmonary malignancies. Whereas cancer-testis antigens are expressed in a variety of human malignancies, immune responses to these proteins are uncommon in thoracic oncology patients due to low level, heterogeneous antigen expression, deficient antigen processing/presentation, and local as well as systemic immunosuppression. Our published studies from cell lines and patient biopsies have demonstrated that thoracic malignancies exhibit diverse patterns of CTA expression and heterogeneous responses to epigenetic regimens that up-regulate CTAs. A strategy to overcome these limitations is to immunize patients against a panel of CTAs that potentially can be up-regulated by systemic administration of chromatin remodeling agents. To address this issue, we conducted a phase 2.5 First-in-Human trial to ascertain if a tumor lysate vaccine can induce broad immunity to CTA and determine if metronomic oral cyclophosphamide and celecoxib (cy/cel) enhances vaccine-induced immune responses. Briefly, 21 patients with primary thoracic malignancies or extra-thoracic neoplasms metastatic to the chest rendered NED by conventional therapies were randomized to receive H1299 lung cancer cell lysates with Iscomatrix adjuvant via deep intradermal injection q month x 6 +/- daily cy/cel. The primary endpoint was serologic response to purified CTA 1 month after the 6th vaccination. Exploratory objectives included analysis of serologic reactivity to carbohydrate antigens and assessment of peripheral immune subsets before and after vaccine therapy. All patients exhibited local and systemic inflammatory responses lasting 72-96 hours following vaccinations. There were no dose limiting treatment related toxicities. 14 patients (67%) completed all six vaccinations. With a median follow-up of 64.6 months (range 58.8-68.4), 12 patients are alive and NED; 2 patients are alive with controlled, second malignancies. 11 of 14 surviving patients (79%) received six vaccinations compared to 3 of 7 patients (42%) who succumbed to their malignancies. Seven of 10 patients randomized to vaccine with cy/cel received all six vaccinations; 3 of these 7 patients (43%) died of disease. Seven of 11 patients randomized to vaccine alone completed six vaccinations; none (0%) of these 7 patients have died. 8 patients (57%) exhibited serologic responses to NY-ESO-1. Additional reactivities were observed against GAGE7, XAGE, and MAGE-C2. Reactivities against tumor-associated carbohydrate antigens were uncommon. Vaccine therapy decreased percent Tregs (p=0.067*), PD-1 expression on Tregs (p=0.023*), and PD-L1 expression on CD14+ monocytes (p=0.0089*), classical monocytes (p=0.0159*), and intermediate monocytes (p= 0.0031*). Cy/cel did not impact immune responses or vaccine-induced alterations in peripheral immune subsets. Laboratory metrics of response did not appear to correlate with patient survival. Results of this positive trial have recently been published online and support further vaccination efforts in patients with thoracic malignancies.