Centre for Mathematics and Physics in the Life Sciences and EXperimental Biology, University College London, London, United Kingdom.
PLoS One. 2013 May 6;8(5):e62320. doi: 10.1371/journal.pone.0062320. Print 2013.
Breast cancer patients have an anomalously high rate of relapse many years--up to 25 years--after apparently curative surgery removed the primary tumour. Disease progression during the intervening years between resection and relapse is poorly understood. There is evidence that the disease persists as dangerous, tiny metastases that remain at a growth restricted, clinically undetectable size until a transforming event restarts growth. This is the starting point for our study, where patients who have metastases that are all tiny and growth-restricted are said to have cancer dormancy. Can long-term follow-up relapse data from breast cancer patients be used to extract knowledge about the progression of the undetected disease? Here, we evaluate whether this is the case by introducing and analysing four simple mathematical models of cancer dormancy. These models extend the common assumption that a random transforming event, such as a mutation, can restart growth of a tiny, growth-restricted metastasis; thereafter, cancer dormancy progresses to detectable metastasis. We find that physiopathological details, such as the number of random transforming events that metastases must undergo to escape from growth restriction, cannot be extracted from relapse data. This result is unsurprising. However, the same analysis suggested a natural question that does have a surprising answer: why are interesting trends in long-term relapse data not more commonly observed? Further, our models indicate that (a) therapies which induce growth restriction among metastases but do not prevent increases in metastases' tumourigenicity may introduce a time post-surgery when more patients are prone to relapse; and (b), if a number of facts about disease progression are first established, how relapse data might be used to estimate clinically relevant variables, such as the likely numbers of undetected growth-restricted metastases. This work is a necessary, early step in building a quantitative mechanistic understanding of cancer dormancy.
乳腺癌患者在明显治愈性手术切除原发肿瘤后多年——长达 25 年——出现异常高的复发率。在切除和复发之间的数年中，疾病进展情况了解甚少。有证据表明，疾病持续存在，以危险的微小转移形式存在，这些转移保持生长受限，临床无法检测到大小，直到发生改变事件重新开始生长。这就是我们研究的起点，在这种情况下，所有微小且生长受限的转移都被认为处于癌症休眠状态。从乳腺癌患者的长期随访复发数据中能否提取出有关未检测到疾病进展的知识？在这里，我们通过引入和分析四种简单的癌症休眠数学模型来评估是否存在这种情况。这些模型扩展了一个常见的假设，即随机转化事件（如突变）可以重新启动微小、生长受限转移的生长；此后，癌症休眠进展为可检测的转移。我们发现，生理病理细节，例如转移必须经历多少个随机转化事件才能逃脱生长限制，无法从复发数据中提取。这个结果并不出人意料。然而，同样的分析提出了一个自然的问题，这个问题有一个令人惊讶的答案：为什么长期复发数据中没有更常见地观察到有趣的趋势？此外，我们的模型表明：(a) 诱导转移中生长受限但不能阻止转移肿瘤形成增加的疗法可能会引入一个手术后时间，使更多患者更容易复发；以及 (b) ，如果首先确定了疾病进展的一些事实，如何使用复发数据来估计临床相关变量，例如未检测到的生长受限转移的可能数量。这项工作是建立癌症休眠定量机制理解的必要的早期步骤。
