40 years of cancer evolution: time to understand its clinical implications

Tumor heterogeneity is observed at many levels: from cancer subclones in primary tumor with distinct characteristics to differences between primary and metastatic tumors in the same individual. As it affects treatment response, it is one of the key challenges in effective cancer management in the clinical practice. Here, co-authors of an article published today in BMC Medicine discuss their study of kidney cancer patients.

40 years ago, Peter Nowell published a visionary model of cancer evolution that describes how inherent genetic instability generates diversity within cancers, enabling the Darwinian selection of mutant subclones that drive tumor progression, metastasis and drug resistance.

Though only a few years ago powerful Next-Generation Sequencing technologies started to enable scientists around the world to assemble the evidence that genetic heterogeneity and cancer evolution are indeed common across most solid tumours types [1]. These data provided a compelling explanation for the frequent failure of novel genotype targeted therapeutics and for the persistently poor outcomes in oncology.

Yet, although intratumour heterogeneity and the fundamental evolutionary nature of cancer are now widely accepted concepts, their impact on daily practice remain negligible where they are thought to have the biggest consequences: in the clinic.

Although intratumour heterogeneity and the fundamental evolutionary nature of cancer are now widely accepted concepts, their impact on daily practice remain negligible where they are thought to have the biggest consequences: in the clinic.

We previously showed that genetic intratumour heterogeneity occurs frequently in clear cell renal cancers [2], the commonest type of kidney cancer. Genetic defects did not only vary within the primary tumor in the kidney, but also between the primary tumor and metastases and between different metastases within the same patient.

Mutations in so-called driver genes, which promote tumor progression and are likely to influence drug sensitivity and resistance, were often heterogeneous. Based on these results, we predicted that treatment responses of different tumor subclones and even of different metastases should vary within individual patients.

In our study published in BMC Medicine, we now tested this hypothesis in 27 patients with kidney cancer who received treatment with anti-angiogenic tyrosine kinase inhibitors (TKIs). To investigate heterogeneity of drug response and resistance, we tracked the size of multiple metastases in each patient.

We analyzed a total of 115 metastases with serial CT scans that had been performed at regular intervals throughout TKI treatment. This showed that the majority of patients (56%) did have a heterogeneous response between metastatic sites.

The biggest surprise came when we analyzed how metastases changed at the time the tumor acquired drug resistance and started to progress again. The occurrence of new metastases defined progression in 67% of patients. At the same time, many of the metastases that had been present from the start of treatment remained controlled in these patients. And in almost half of these patients, the combined size of metastases that was still controlled was larger than that of the progressing ones.

The biggest surprise came when we analyzed how metastases changed at the time the tumor acquired drug resistance and started to progress again.

These results show that drug response and resistance heterogeneity are common in clear cell kidney cancers. Unfortunately, tumor samples to compare genetic heterogeneity with the radiological results were not available. Nevertheless, these results support our initial hypothesis and mirror the intratumor heterogeneity which we found at the genetic level in our previous work.

These new data is likely to have important implications for treatment decision making. For example, continuation of TKI treatment may be beneficial in patients in whom the bulk of the metastases remain controlled. Individual progressing lesions may benefit from focal therapies.

In addition, radiological criteria that evaluate drug treatment efficacy (so called ‘Response Evaluation Criteria in Solid Tumors’ or RECIST criteria) do not capture or report heterogeneity. Treatment switches or discontinuation may hence occur unnecessarily early in patients with small volume progression while the disease bulk remains under control. Potential differences in prognosis and probability of benefit from second line therapies between patients with homogenous and heterogeneous progression patterns remain unknown as a consequence.

Analyses of drug response heterogeneity may also help to evaluate new therapies and define optimal combination treatments in metastatic cancers. New immunotherapies have already changed the outcome of cancers such as melanoma, where long term responses can now be achieved in a significant proportion of patients. Checkpoint blocking immunotherapy also has activity in renal cell carcinomas.

It will be important to assess if the distinct ability of the immune system to evolve and to potentially recognize and target multiple heterogeneous subclones reduces heterogeneous responses.

It will be important to assess if the distinct ability of the immune system to evolve and to potentially recognize and target multiple heterogeneous subclones reduces heterogeneous responses. Also, combination therapies that suppress heterogeneous responses could lead to longer disease control than those frequently leading to heterogeneous responses. Such data could be extracted even from small, early stage clinical trials by systematically investigating radiological heterogeneity.

In order to change oncological practice, heterogeneity studies using radiological and genetic approaches, for example based on liquid biopsy approaches that require simple blood samples [3], should be routinely incorporated into clinical trials. That would require establishing better radiological response criteria, which incorporate heterogeneity measures and quantify the proportion of the cancer that is progressing.

Such approaches could rapidly assemble the data that helps to define how intratumor heterogeneity influences clinical outcomes and to provide the rational for the development of effective, heterogeneity-aware interventions. This could eventually change clinical practice, so that instead of treating cancers as a uniform and static disease, we treat them as evolving entities.


[1] Gerlinger M et al. Cancer: evolution within a lifetime. Annu Rev Genet. 2014(28):215-36

[2] Gerlinger M et al. Intratumour heterogeneity and branched evolution revealed by multiregion sequencing. N Eng J Med. 2012 366(10):883-92

Gerlinger, M et al. Genomic architecture and evolution of clear cell renal cell carcinomas defined by multiregion sequencing. Nat Genet. 2015 46(3):225-33

Gulati, S et al. Systematic evaluation of the prognostic impact and intratumour heterogeneity of clear cell renal cell carcinoma biomarkers. Eur Urol 2014 66(5) 936-48

[3] Bettergowda, C et al. Detection of Circulating Tumor DNA in Early- and Late-Stage Human Malignancies. 2014 19;6(224)

View the latest posts on the On Medicine homepage

Comments