The results of an initiative designed to better understand the origins of human cancers, known as the National Institutes of Health-sponsored Pan-Cancer Atlas, were published in a series of 27 articles in the March and April 2018 issues of the Cell Press journals.

One of the most profound implications of the findings is that factors other than cell and tissue origin of tumours, such as molecular similarities between cancers arising at different anatomical sites, should be taken into account upon classification and subsequent treatment. Given that solid tumours and haematological malignancies have traditionally been treated according to location, this could eventually signal a major paradigm shift in drug development, the design of clinical trials, and patient treatment algorithms, thereby requiring re-examination and possible revamping of current cancer treatment guidelines.

The project studied 33 of the most prevalent cancer types, and found that while these cancers fell into four major cell-of-origin patterns, molecular clustering analyses revealed 28 patterns in which these cancers could be grouped. Of note, one-third of the patterns contained tumours of the same classification type, while the other two-thirds of the groups each contained cancers that were more heterogeneous, although most groups were limited to specific cell and tissue patterns.

Varying treatment guidelines

Treatment guidelines for certain tumour types can vary according to cell of origin of a tumour (for example, squamous vs nonsquamous), but are generally tissue type-specific, so squamous cell carcinomas of the lung are treated differently from head and neck squamous cell carcinomas. The molecular subgroupings, however, reveal similarities in cancers arising from particular cell types across different tissues, implying at the very least that a re-examination of current treatment guidelines would be useful.

Given the implications of such a dramatic shift in cancer classification from tissue location to that of molecular determinants, a more conservative approach could simply be to recognise and invest more in the validity of tumour-agnostic approaches in current treatment paradigms.

This is already occurring to some extent, heralded by the May 2017 FDA accelerated approval of the checkpoint inhibitor Keytruda for any non-resectable, metastatic solid tumors that are positive for either the microsatellite instability-high (MSI-H) or mismatch repair deficient (dMMR) biomarkers and that have progressed on previous treatment or for which there are no satisfactory alternative treatment options.

While this approach may not work for all tumours expressing particular biomarkers, such as BRAF and MEK mutations, which respond to targeted drugs in melanoma and lung cancers, but not colorectal cancers, there are some promising signals for further success for biomarker-driven, tumour-agnostic approaches. Bayer and Loxo Oncology completed a rolling NDA submission to the FDA in March 2018 for larotrectinib, a tropomyosin receptor kinase (TRK) inhibitor, to treat all locally advanced or metastatic solid tumours that possess neurotrophic tyrosine receptor kinase (NTRK) gene fusions.

Interestingly, the results of several trials showed efficacy as well as a good safety profile across nearly 20 pediatric and adult cancer types, indicating that a similar approach in the case of other biomarkers is worthwhile pursuing. Now that the results of the Pan-Cancer Atlas have been released, molecular similarities between tumour types can be used to inform biomarker-driven clinical trials across different tumour types sharing the same determinants, facilitating the development of effective tumour-agnostic therapies.

For more insight and data, visit the GlobalData Report Store – Pharmaceutical Technology is part of GlobalData Plc.