Clinical, Basic, Translational and Kinetic Studies of Drug Action and Resistance

我们的长期目标是尽可能深入地了解药物疗效的机制。经验告诉我们，不仅对于我们的传统细胞毒剂，而且对于我们的新的“靶向治疗”，我们的药物未能使患者受益的主要原因是耐药性的产生。固有的和获得性的耐药性都是成功结果的主要障碍。我们坚信，尽管抗药性可能是复杂的，但它不是一个无法克服的问题。我们还坚信，存在的机制数量有限，特别是在分子水平上，我们对这些机制了解得越好，我们就越有可能开发出有效的治疗方法。我们相信，从我们的耐药性模型中吸取的经验教训--无论是临床前还是临床--将广泛适用于其他药物。因此，利用我们对我们研究的机制的深入了解，我们有信心获得具有广泛适用性的知识。我们并不只关注一种抵抗机制。我们考察了抗性机制的全部内容。我们越来越多地通过检查大量的临床试验数据并对肿瘤生长和退化的动力学进行详细、深刻和深入的分析来做到这一点。我们已经开发了一种新的范例，用于使用患者在临床试验中获得的肿瘤测量来评估治疗效果。这个数学模型适用于许多肿瘤类型，并可能有助于评估结果。使用在患者参加临床试验时收集的数据来模拟肿瘤进展在药物开发和初级肿瘤学护理中可能是有价值的。我们基于这样的模型建立了一个方程，即肿瘤大小呈指数下降(即，作为一个一级过程)，但肿瘤也存在独立的指数再生长，这些反映在血清标记物的数量或成像测量中。这个公式是：f=exp(减去dxt)+exp(Gxt)减1，其中exp是自然对数的底，e=2.7182...，f是时间t的肿瘤量，归一化到开始治疗的第0天的值。速率常数d(衰变，以天数增加到-1)表示肿瘤数量的指数下降，而速率常数g(生长，也以天数增加到-1)表示治疗后肿瘤的指数再生长。我们现在有大量的数据表明，在前列腺癌、肾细胞癌、乳腺癌和多发性骨髓瘤中，生长速率常数(G)与总存活率有非常好的相关性，而回归速率常数(D)则不是。这并不出人意料，因为死亡不是由退化的肿瘤部分引起的，而是由存活和生长的部分以及它生长的速度引起的。我们正在进行的分析旨在明确地证实这一点，以便我们可以提出生长速率常数(G)作为有价值的临床试验终点。但我们也在进一步发展，开发新的范例来评估临床试验数据，到目前为止，这些数据只被描述为三个简单的终点：无进展生存率、总生存率和应答率。尽管收集了大量数据，但这些是我们定义的唯一端点。我们的数据分析旨在扩大这一范围，以便我们可以更好地了解药物是如何起作用的，不同的组合如何比较，什么让一种疗法更好，哪种疗法在佐剂或新佐剂环境中更有可能成功，哪种疗法更有可能失败，以及无数其他分析和相关性。我们还计划使用这些数据来调查和询问迄今为止仅限于临床前模型的假说。实际上，我们将使用终极模型中的终极实验--患有癌症的人类--来理解基本的生物学原理。这一部分将描述已提交的、正在进行的和拟议的研究，以了解临床上耐药的机制，并更好地定义我们的治疗方法是如何工作的。我们的研究目标是：(1)了解耐药的分子基础；(2)了解这些变化是如何/为什么发生的；(3)设计减少或防止其发生的策略；(4)验证和开发用于评估治疗效果的新范式。在临床上，贝茨博士和福霍博士继续进行耐药癌症的试验。这一部分代表了我作为肿瘤学领域的研究人员开始职业生涯的地方。这也反映出我认识到，在这个资源定义明确的时代，并不是所有想要调查的研究问题都可以继续下去。因此，它不建议进行实验室实验。相反，它寻求利用我认为只是表面挖掘的海量数据。由于许多大型随机临床试验招募了数百名患者并从每个试验中获取了大量数据，值得注意的是，所有这些数据的输出通常是三个数字：(1)总体应答率；(2)中位无进展存活率；(3)中位总存活率。在我们幸运地获得的几个试验中，有一万到一万五千个数据点。如果一位基础科学家用750只小鼠进行活体实验，收集了临床试验中定期收集的所有数据，并且他/她的分析只涉及三个终点，我们会怎么想？正如我们正在发现的那样，有大量的数据需要挖掘，它可以为我们提供许多多年来一直试图回答的问题的洞察力--在真实的患者中！

Our long-term goals are to understand in as much depth as possible the mechanisms of drug efficacy. Experience has taught us that not only for our traditional cytotoxic agents, but also for our new "targeted therapies" the major reason why our drugs fail to benefit patients is the development of drug resistance. Both intrinsic and acquired drug resistance are the major impediments to a successful outcome. We firmly believe that while drug resistance can be complex it is not an insurmountable problem. We also firmly believe that a limited number of mechanisms exist, especially at a molecular level, and that the better we understand these the more likely we are to develop effective therapies. We believe that the lessons learned in our models of drug resistance - be they pre-clinical or clinical - will have broad applicability to other drugs. Thus exploiting our in depth understanding of the mechanisms we study we are confident that we will gain knowledge with broad applicability. We do not focus solely on one mechanism of resistance. We examine the full breadth of resistance mechanisms. Increasingly we do this by examining large clinical trial data and conducting detailed, incisive and in-depth analysis of the kinetics of tumor growth and regression We have developed a novel paradigm for assessing therapeutic efficacy using tumor measurements obtained while patients are enrolled in a clinical trial. This mathematical model has applications to many tumor types and may aid in evaluating outcomes. Modeling tumor progression using data gathered while patients are enrolled on a clinical trial could be valuable in drug development and in primary oncology care. We developed an equation based on the model that tumor size decreases exponentially (i.e., as a first-order process) but that there is also independent exponential re-growth of the tumor and these are reflected in the quantity of a serum marker or imaging measurements. This equation is: f = exp(minus d x t) + exp(g x t) minus 1 where exp is the base of the natural logarithm, e = 2.7182 ..., and f is the tumor quantity at time t, normalized to the value at day 0, the time at which treatment is commenced. The rate constant d (decay, in days raised to the minus 1) accounts for the exponential decrease in the tumor quantity, whereas the rate constant g (growth, also in days raised to the minus 1) represents the exponential re-growth of the tumor following treatment. We now have extensive data that in prostate cancer, RCCs, breast cancer and multiple myeloma that show the growth rate constant (g) correlates exceptionally well with overall survival while the regression rate constant (d) does not. This is not unexpected since death is not caused by the fraction of tumor that regresses, but by the fraction that survives and grows and how fast it grows. Our ongoing analyses are designed to confirm this unequivocally so that we may propose the growth rate constant (g) as a valuable clinical trial endpoint. But we are also developing the further by developing novel paradigms to assess clinical trial data that heretofore has only be presented as three simple endpoints: progression-free survival, overall survival and response rate. Despite collecting a large amount of data these are the only endpoints we define. Our data analysis looks to expand this so that we may better understand how drugs work, how different combinations compare, what makes one therapy better, which therapy might be more likely to succeed in an adjuvant or neo-adjuvant setting and which one more likely to fail and numerous other analyses and correlations. We also plan to use the data to investigate and interrogate hypotheses heretofore confined to pre-clinical models. In effect we will use the ultimate experiment in the ultimate model, humans with cancer, to understand basic biologic principles. This section will describe submitted, ongoing and proposed studies to understand the mechanisms of drug resistance clinically and to better define how our therapies work. Our research goals are to (1) understand the molecular basis of drug resistance; (2) comprehend how/why these changes occur; (3) devise strategies to reduce or prevent their occurrence and (4) validate and develop a novel paradigm for assessing therapeutic efficacy. In the clinic, Drs. Bates and Fojo continue to conduct trials in drug-resistant cancers. This section represents where I started my career as a researcher in the field of oncology. It also reflects an acknowledgement on my part that in this age of clearly defined resources, not all research questions that one would like to investigate can be pursued. So it proposes no laboratory experiments. Rather it seeks to leverage what I believe is an enormous amount of data that is mined only superficially. With many large randomized clinical trials enrolling hundreds of patients and capturing an enormous amount of data from each one, it is remarkable that the output from all of this data is usually three numbers: (1) overall response rate; (2) median progression-free survival; and (3) median overall survival. Several of the trials that we have been lucky to obtain have ten to fifteen thousand data points. What would we think if a basic scientist conducted an in vivo experiment with 750 mice, collected all the data regularly collected in a clinical trial and his/her the analysis referred to only three endpoints? As we are discovering there is a trove of data to be mined and it can provide us insights into many of the questions we have been trying to answer for many years - in actual patients!