Carboplatin Response Rate Not Superior to Docetaxel in BRCA1 Methylated or Silenced TNBC

Andrew Tutt, MD

Testing for BRCA1 methylation or mRNA silencing did not predict a better response to carboplatin over docetaxel in patients with advanced triple-negative (TNBC) or BRCA1/2 breast cancer, according to results presented at the 2016 San Antonio Breast Cancer Symposium.

The findings come from a pre-planned biological analysis of the phase III TNT trial, a study in which patients with TNBC and germline BRCA1/2 mutations previously achieved a higher response rate with carboplatin than docetaxel. However, these latest results actually demonstrated that patients with tumors with BRCA1 methylation had a higher response when treated with docetaxel than with carboplatin (42.1% vs. 21.4%; P = .28).

Thus, these results do not support testing for BRCA1 methylation or mRNA silencing to guide treatment in metastatic TNBC, according to Andrew Tutt, MD, consulting clinical oncologist at Guy’s Hospital Breast Unit, and professor of Oncology at King’s College.

Patients enrolled in the TNT trial had estrogen receptor-negative, progesterone-receptor-negative, HER2-negative breast cancer or had a known germline BRCA1/2 mutation. They were randomized to carboplatin or docetaxel for 6 to 8 cycles or until disease progression, at which time crossover to the alternate therapy was appropriate.

The a priori subgroup analyses were performed around “BRCAness” populations, said Tutt, including those with germline BRCA1/2 mutations and those with mutational signatures of homologous recombination (HR) deficiency, including patients with epigenetic dysregulation of BRCA1 through methylation and silencing of the BRCA1 mRNA.

“In cancer, aberrant methylation of cytosines frequently occurs in the context of CpG dinucleotides in the regulatory regions of genes,” explained Tutt. “This can be associated with genetic silencing of that gene’s transcription with functional consequences. The regulatory region of BRCA1 is known to be subject to such epigenetic silencing.”

Methylation in the regulatory region of BRCA1 occurs in about 10% to 40% of TNBCs. CpG methylation above the 10% level is associated with silencing of BRCA1 mRNA.

The a priori hypothesis was that BRCA1 methylation would show a significant interaction with carboplatin treatment effect, as was done for the germline BRCA1/2 population.

DNA from 224 primary tumors was subjected to BRCA1 methylation status, and a methylation percentage for the sample was computed as the proportion of methylated reads relative to the total unambiguous reads. A sample with methylation >10% was considered methylated (pre-specified). Results from 212 of the samples met quality control; 33 of these samples (18%) were considered to be BRCA1 methylated and 179 were non-methylated. BRCA1 methylation and germline BRCA1 mutation co-existed in 2 patients.

“Counter to our hypothesis, those patients with BRCA1 methylation had a higher response to docetaxel than to carboplatin,” said Tutt. The objective response rates (ORRs) were 21.4% in the BRCA1 methylated patients randomized to carboplatin compared with 42.1% in those randomized to docetaxel. The difference did not achieve significance (P = .28).

In patients without BRCA1 methylation, ORRs between the 2 agents were similar: 34.4% in the carboplatin group and 38.4% in the docetaxel group (P = .64).

The test for interaction for response by BRCA1 methylation status did not achieve significance (P = .35).

BRCA1 methylation had a different relationship with treatment response to germline BRCA mutation. Response rates in the 31 patients with BRCA methylated and germline BRCA wild type were 15.4% in the carboplatin group versus 44.4% in the docetaxel group (P = .13). In the 43 patients with germline BRCA mutations, response rates were 68.0% in those assigned to carboplatin and 33.3% in patients assigned to docetaxel (P = .03).

Some 218 tumor samples were tested for BRCA1 mRNA silencing status, and results were available for 191. A cutpoint of value of log₂(mRNA)<8.4 was established for BRCA mRNA low status using a bimodality analysis. Thirty-one (16%) were found to be silenced by this definition. In 184 patients, both mRNA status and methylation status were available. Nineteen of 29 methylated samples (66%) were defined as silenced.

Consistent with the results by BRCA methylation, “we again showed, counter to our hypothesis, that the BRCA1 silenced population had a higher response rate to the standard-of-care drug docetaxel than to carboplatin,” said Tutt. The difference did not achieve significance (P = .073). In the non-silenced population, the 2 agents produced similar response rates.

The test for interaction was negative (P = .066) and exclusion of the BRCA-mutated population did not change the overall result.

Unlike with germline BRCA1/2 status, there was no significant effect of BRCA methylation on PFS following either carboplatin or docetaxel.

Exploratory analyses examining potential relationships between a response to carboplatin and new cutpoints for BRCA1 methylation or BRCA1 mRNA level did not suggest evidence of a signal.

Tutt proposed a hypothetical model to explain the differences in treatment interactions. In a primary tumor with high heterogeneity with many clones non-methylated within the tumor, and under the selective pressure of DNA-damaging chemotherapy, residual micrometastases frequently lose methylation or have been selected for a non-methylation status. By losing its methylation and its HR defect, it will not respond to platinum chemotherapy. In contrast, primary tumors mutated for BRCA1/2 have a high degree of homogeneity for mutation and functional loss through loss of the wild-type allele. The germline-mutated tumor retains its BRCA mutation in the micrometastasis, and at relapse, retains its mutation and HR defect, rendering it more sensitive to platinum chemotherapy.

Tutt A, Cheang MCU, Kilburn L, et al. BRCA1 methylation status, silencing and treatment effect in the TNT trial: A randomized phase III trial of carboplatin compared with docetaxel for patients with metastatic or recurrent locally advanced triple negative or BRCA1/2 breast cancer (CRUK/07/012). Presented at: 2016 San Antonio Breast Cancer Symposium; December 6-10, 2016; San Antonio, TX. Abstract S6-01.