NEW YORK – Based on a gene expression score, a subset of HR+/HER2- breast cancer patients appears more likely to respond to the histone deacetylase 6 inhibitor ricolinostat, according to a recent study.
The study, published last week in Nature Cancer, applied a computationally derived gene expression measure of HDAC6 function, termed an HDAC6 score, to tumor specimens obtained from 16 patients with any metastatic breast cancer subtype, who were taking ricolinostat and nab-paclitaxel.
The two compounds were safe and well tolerated in the study, with clinical activity seen in patients with HR+/HER2− disease, and the HDAC6 score showed potential as a predictive biomarker for response to ricolinostat.
HDAC6 function is essential for an aggressive breast cancer subtype called inflammatory breast cancer (IBC), which accounts for an estimated 1 to 4 percent of all breast cancers. Although HDAC6 is rarely amplified or mutated in breast cancers, its protein activity, as measured by the expression levels of genes that it regulates (its regulon), changes markedly, appearing to grow overactive in IBC.
Jiyang Yu, the study's principal investigator, and his colleagues at St. Jude Children's Research Hospital and Columbia University reasoned that along with IBC, other breast cancers might also show a dependency on HDAC6 and could therefore be susceptible to HDAC6 inhibitor therapy.
By applying the HDAC6 score to all primary and metastatic tumors included in datasets from The Cancer Genome Atlas (TCGA), the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC), and the Metastatic Breast Cancer (MBC) project, the team identified a group comprising approximately 30 percent of all breast cancers whose score suggested a response to HDAC6 inhibition and were enriched in hormone receptor-positive tumor types.
They confirmed the correlation between the HDAC6 score and anticancer response to HDAC6 inhibition with ricolinostat in 14 breast cancer cell lines and in mouse models before launching a Phase Ib dose escalation study of ricolinostat with nab-paclitaxel, in partnership with Acetylon/Celgene (NCT02632071).
They included paclitaxel based on preclinical data showing synergistic activity with ricolinostat, as well as for consistency with standard of care breast cancer treatment.
They found that the combination was safe and well-tolerated, with clinical activity evident specifically in the 13 patients with HR+/HER2- disease and three with triple-negative metastatic breast cancer. A subsequent retrospective analysis that included DarwinHealth's bioinformatic OncoTarget platform, which assays for aberrant and pharmacologically actionable proteins in patient tumor samples, showed that the HDAC6 score could stratify patients based on clinical outcomes.
Further analysis showed that the anticancer activity of ricolinostat was associated with reduced expression of c-Myc, a substrate of HDAC6, that is degraded during HDAC6 inhibition.
The study "identifies a potential biomarker for identifying subtypes of tumors that may be sensitive to HDAC6 inhibition," said Triona Ni Chonghaile, a senior lecturer at the Royal College of Surgeons in Ireland, who researches targeted breast cancer therapies.
One aspect of the study that she found interesting was that triple-negative breast cancers, which are thought to be driven by Myc, were not as sensitive as HR+HER2- tumors in the study.
She also wondered whether the biomarker signature only works for ricolinostat or also for other HDAC6 inhibitors.
Yu responded that his group tested "quite a few" HDAC6 inhibitors and that the signature does not appear to be specific to ricolinostat, although that work is at an earlier stage than the published results.
"We're testing another combination to target the other pathways that seems to be active upon HDAC6 inhibition," Yu said.
Chonghaile also commented that the reason why HDAC6 inhibition appears to acetylate and degrade Myc in selective lines remains an open question, although Yu suggested that these cell lines are likely to simply be those that depend on HDAC6 activity.
Although Yu says that the study provides the first evidence that the HDAC6 score has predictive potential in a clinical setting, he acknowledged that it needs further refinement.
"Triple-negative patients also responded," he said, "[which] means we definitely need a better biomarker for the use of HDAC6 inhibitors."
Yu and his colleagues plan to conduct other Phase I and II clinical studies of their HDAC6 score in both breast cancer and other cancers, and using other HDAC6 inhibitors. Although Celgene helped to sponsor the trial, Yu said that the company has no further involvement with his lab, as its main focus is on immunotherapies.
"We are reaching out to a few pharmas who have HDAC6 inhibitors and a few biotechs who can help develop a clinical-grade kit to report HDAC6 score," he said. "We will also expand the HDAC6 score to other cancer types, [such as] hematologic cancers, and neurological diseases, and seek support to evaluate context-specific HDAC6 scores for prediction of response to HDAC6 inhibitors."
Eventually, Yu hopes to seek companion diagnostic clearance from the US Food and Drug Administration.
"Our current HDAC6 score is based on total RNA-seq profile, which might be challenging to obtain FDA approval," he said. "We are optimizing the HDAC6 regulon for score calculation and seeking support to develop a test kit, [such as a] targeted RNA-sequencing, or PAM50-like kit that [the] FDA accepts."