Large list of potentially targetable genes, but what about outcomes?
Interest is great in genomically informed targeted therapy, with the goal of identifying genomic alterations that (1) are drivers of tumor growth and progression in individual patients to individualize therapy; and (2) are targetable directly or indirectly with approved or investigational agents.
But should all women diagnosed with metastatic breast cancer undergo next-generation sequencing (NGS)? The question was debated by two experts at the most recent San Antonio Breast Cancer Symposium.
Yes, said Funda Meric-Bernstam, MD, chair of Breast Cancer Research at the University of Texas MD Anderson Cancer Center in Houston. Genomic testing should be part of the clinical management, and should be considered in all patients with metastatic breast cancer and adequate performance status who have an interest in clinical trials.
A large list of genes are potentially targetable in breast cancer, she said, pointing to PIK3CA, Akt, HER2, TRK and other rare alterations.
Several PI3K inhibitors are in clinical trials to target PIK3CA, with “emerging hope in upcoming inhibitors such as alpelisib in combination with fulvestrant” in PIK3CA-altered advanced breast cancer, she said. Activating Akt mutations, usually E17K, are most commonly found in hormone receptor-positive breast cancer. Objective responses have been elicited with the catalytic inhibitor AZD5363 as monotherapy in patients with estrogen receptor-positive AktE17K-mutant breast cancer and is now being studied in combination with fulvestrant. Ipatasertib, another Akt inhibitor, combined with paclitaxel in patients with PIK3 pathway aberrations increased progression-free survival to 9.0 months, compared with 4.9 months with paclitaxel alone in a phase II study.
HER2 is a proven genetic target, Meric-Bernstam said, noting that some patients who are HER2-negative on initial screening are subsequently found to be HER2-positive on NGS of another or newer sample. “We’re not sure if this is genomic evolution or heterogeneity or technical issues with the first testing, and we’re not as sure of the therapeutic sensitivity in this context, especially if it represents heterogeneity.”
If the tumor is HER2 amplified on NGS, validation is not needed to institute HER2-directed therapy. If the tumor is not amplified on NGS, the patient may still have a lower level of amplification or overexpression. “There’s a lot of enthusiasm about exploring HER2 mutations as a target.”
A few years ago, activating HER2 mutations were discovered in HER2-negative breast cancer. In a series of 5,605 women with breast cancer who underwent genomic profiling, 10.6% had HER2 amplifications, 2.4% had HER2 mutations, and 0.7% had co-occurring HER2 amplification and mutation, she continued. A few agents have already been approved in the HER space, with neratinib being the most prominent.
A very rare alteration found in several tumor types including breast cancer is TRK fusions. As presented at the 2017 ASCO annual meeting, in a phase I/II basket trial of patients with TRK (tropomyosin receptor kinase) fusions, almost all patients treated with the pan-TRK inhibitor larotrectinib had an objective response, which proved durable. Meric-Bernstam explained that TRK fusions are pathognomonic in secretory breast cancer, which constitutes less than 1% of breast cancers. “Because they are rare in breast cancer, I am not going to advocate for TRK fusion testing across the board for this reason, but if you do have a secretory breast cancer patient, please do TRK fusion testing.”
Not performing NGS means that patients with rare alterations cannot be entered into genotype-selected clinical trials, she argued.
ESR1, another current clinically relevant mutation, is rarely found in primary breast cancer but is commonly found in the metastatic setting. Evidence suggests that as ESR1 mutations accumulate with further treatment, there may be some value in retesting or performing liquid biopsy. An ESR1 mutation may affect the choice of therapy; an improved PFS was obtained with the use of fulvestrant compared with exemestane in breast cancer patients with ESR1 mutations, which was not the case in the patient who were ESR1 wild type.
Outcomes Not Altered, Potential Pitfalls Remain
The debater taking the other side at the symposium, Fabrice André, MD, of Gustave Roussy Cancer Center in Villejuif, France, said the key question is whether in the context of routine practice, NGS should be considered for detection of somatic mutations. At least as of yet, he said, no such rationale exists.
At present, no drug approved for use in the treatment of breast cancer requires a genomic test, he reminded listeners. “The reason is because the current way of interpreting DNA sequencing is not useful in metastatic breast cancers, and is potentially deleterious.”
Although NGS has been able to detect alterations in PIK3CA, Atk1, ERBB2, and ESR1 for which objective responses have been observed with the use of targeted therapy in early phase study, their detection did not improve outcome. Progression-free survival in these studies ranged from 5 to 8 months, which was not superior to standard of care. Further, these alterations can be detected by polymerase chain reaction (PCR) assays on circulating tumor DNA, which would be less expensive than NGS, he said.
In the case of sensitivity to PD-1 inhibitors such as pembrolizumab, accelerated approval of the agent was granted in those patients with microsatellite instability-high or mismatch repair-deficient solid tumors. The companion diagnostic for this purpose is immunohistochemistry or PCR, not NGS, said André. “Keep in mind that breast cancer with microsatellite instability is extremely rare — something like 1% and mostly in triple-negative breast cancer.”
The largest commercially available NGS panel can detect about 300 targetable genomic alterations. The issue here is that large gene panels report targetable alterations that are not relevant or for which the wrong drug may be recommended. Therefore, NGS reports can be deleterious because they recommend ineffective therapy and deny effective therapies, he said — one illustration of the wrong target, for example, is fibroblast growth factor receptor (FGFR)1/2 amplification for which an FGFR may be recommended. Nogova et al reported an objective response rate of 0% in patients with breast cancer and FGFR1/2 amplification.
Two large clinical trials in which large gene panels were used had efficacy as the primary endpoint, and in both cases, the targeted drugs matched to the genomic alteration detected by NGS failed to improve PFS.
Finally, said André, the reporting of large panels of genes leads to major ethical, regulatory, and financial issues that have not yet been sorted out. For example, in comparing results obtained from different NGS vendors, the overlap in genomic alterations is sometimes poor. Another pitfall is that somatic genetic testing in patients with advanced cancer may also detect previously unrecognized pathogenic germline variants.
Furthermore, he said, with genomic testing the likelihood of finding a drug matched to a genomic alteration is low. Although sequencing is inexpensive, it generates additional cost for biopsies and potentially the off-label use of expensive drugs.