A caveat to basket trials is that rearrangements may not exhibit the predicted response to drugs. Such an example is the BRAF fusion in comparison to the BRAF V600E point mutation. It is established that melanomas, pilocytic astrocytomas, and lung cancers with BRAF V600E mutations are sensitive to BRAF inhibitors, eg, vemurafenib. However, paradoxical activation can occur in tumors with BRAF fusions. In patients with these fusions, alternative inhibitors, eg, sorafenib, which inhibits feedback signaling through CRAF, or trametinib, which inhibits downstream MEK, may serve as better options in inhibiting tumor growth.77,116,117 Thus, fusions may confer differential functionality as compared to activating point mutations in terms of the mechanism of signaling and sensitivity to inhibition.
Although the prevalence of fusions in breast cancer has been reported in many studies,3,34,36 much less is known about the role of fusions in breast cancer. Functional evaluation of such genes is critical to understand the scope of therapeutic relevance. Validation of top candidates found in cancer cell lines and tumor tissue cohorts suggest potential oncogenic mechanisms and therapeutic opportunities. Table 4summarizes some of the fusions reported to be recurrent in breast cancer cell lines and tumor subtypes, their oncogenic characterization, prevalence in other cancers, and potential therapeutic opportunities for each. In the latest precision medicine approach for difficult-to-treat cancers, regardless of the recurrence of fusion genes, even the presence of one or more actionable kinase fusions is applicable for targeted therapy.
Small molecule kinase inhibitors are the most commonly used targeted approach (Table 3). However, kinase fusion partners may give rise to their own functional consequences and have also been exploited for the development of targeted treatment strategies. Use of ABL kinase inhibitors for BCR-ABL fusion-positive CML is a successful strategy, but drug resistance may be encountered due to point mutations in the kinase domain. Thus, an alternative is targeting the coil–coil dimerization domain located within the N terminus fusion partner BCR. It is believed that this dimerization domain drives kinase activation. Its targeting has been attempted in cell lines.118–121 The specificity and success of the strategy are yet to be evaluated in vivo. Since many kinase fusion partners are found to have coiled coil domains,3,74,122 eg, EML4 of EML4-ALK; ERC1 of ERC1-RET seen in breast cancer,3 this is a potentially important strategy, given the development of resistance to kinase inhibitors during tumor evolution.
Similar to kinase fusions, transcription factor fusions have also been explored for targeted therapy. For the acute promyelocytic leukemia-associated PML-RARA fusion, arsenic trioxide and all-trans retinoic acid are currently used to target PML, the 5′ moiety, and RARA, the 3′ moiety of the fusion protein, respectively. The precise molecular mechanism of action in both cases involves degradation of the PML-RARA oncoprotein, albeit by different pathways.123,124 Interestingly, the mechanism was elucidated only long after the identification of their therapeutic efficacy. Several groups have reported many transcription factor fusions in breast cancer cell lines90,91,93 and tumors, which suggest their widespread involvement in fusion genes with potential functional implications in breast cancer.
Detection of relevant and actionable genomic alterations is at the forefront of personalized therapeutic practice in cancer. Apart from therapy selection and prognosis, there have been clinical reports indicating the usefulness of rearrangements in providing diagnostic clarity,125,126 investigating mechanisms of acquired drug resistance,127 and exploring novel combinatorial therapy. High-throughput genomic sequencing studies, such as exome sequencing of human cancer by TCGA and use of limited hotspot panels, have focused on identifying point mutations and rearrangements specifically involving exons. However, intronic rearrangements are also common in many solid tumors, including breast cancer, and may represent a large class of actionable genomic alterations that are missed by standard short-read sequencing approaches. Inexpensive, rapid turnaround, and clinically implementable sequencing approaches that readily identify potentially actionable genomic rearrangements are clearly needed in conjunction with continued characterization of novel fusion genes.