NEW YORK (GenomeWeb) – A team led by researchers at the Spanish National Cancer Research Center and Utrecht University have identified a set of protein kinases that are predictive of treatment outcomes in triple-negative breast cancer.
Published today in Nature Communications, the study used a combination of mass spec-based phosphoproteomics and immunohistochemistry to identify six kinases whose activation levels can help predict the likelihood of relapse and guide therapy in TNBC.
As the authors noted, TNCB is a difficult-to-treat form of breast cancer with a typically poor prognosis and a lack of good prognostic and predictive markers. The disease is highly heterogeneous at the genetic level, they added, with previous studies identifying "sets of mutations unique to individual patients" and "highly penetrant oncogenes … rarely found."
This, they wrote, suggests that "the TNBC phenotype is a result of coexisting, moderately penetrant, genetic changes that together contribute to its clinical presentation."
This sort of genetic landscape makes it difficult to identify broadly useful markers at the nucleic acid level. However, the researchers hypothesized that the various TNBC-linked genetic alterations might "coalesce into a discrete number of phosphorylation-driven patterns of activation of the proteome, the activity of which would determine the prognosis of a TNBC patient."
If, indeed, genetic aberrations in TNBC were integrated at the level of protein signaling by a relatively small number of dysregulated kinases, these proteins could prove useful not only as prognostic biomarkers for the disease, but also as potential drug targets.
To explore this hypothesis, the researchers used mass spec to analyze the phosphoproteome of 34 tumor samples taken from patients with TNBC, selecting patients who either relapsed within three years of their initial treatment or who were free from relapse after 12 years or more. Their analysis identified more than 10,000 unique phosphosites that mapped to 2,643 different proteins and identified 159 peptides that had significantly higher phosphorylation levels in relapsed patients than in non-relapsed patients.
Using kinase set enrichment analysis (KSEAS), the researchers then identified the kinases responsible for these increased phosphorylation levels, arriving at 11 hyperactive kinases that appeared linked to relapse in TNBC.
They then developed immunohistochemistry assays to these 11 kinases, and in a set of 113 TNBC tumor samples they found that six of the 11 kinases were predictive of relapse when analyzed via IHC. In the 81 of the 113 patients with hyperactivity in one or more of the six kinases, relapse rates were 47 percent after 12-plus years. In the 32 cases with hypoactivity in any one of the six kinases, relapse rates were 6.5 percent.
The authors also compared the observed kinase activation patterns to the mutations present in the different samples, they wrote, finding "that each kinase activation pattern could be achieved by different mutational landscapes," which highlights "the importance of a kinase-based classification as compared with one based on genes."
The researchers then investigated whether patients' kinase profiles might help guide therapy, using the kinase data to inform testing of six kinase inhibitors in doublets across a set of TNBC cell lines, mouse models, and patient derived xenografts. This effort found that kinase levels "appear to aid selection of active and relatively specific doublets in TNBC preclinical models," the authors wrote, though they noted that "future mechanistic insights into the manner by which the activity of the [six] kinases drives TNBC progression … are necessary to ascertain how and when to use clinical grade inhibitors."
Based on the results, the authors wrote that they believed "future studies and trials by independent groups should now find an adequate place for tumor phosphoproteomics in routine clinics, akin to the case of the seminal studies performed with gene expression almost 20 years ago."