Carboplatin Does Not Produce Greater Response Over Docetaxel in BRCA1 Methylated or Silenced TNBC

December 13, 2016
Wayne Kuznar

BRCA1 methylation or silencing does not predict for a better response to carboplatin over docetaxel in women with advanced triple-negative or BRCA1/2 breast cancer, according to results from a pre-planned biological analysis of the phase III TNT trial.

Andrew Tutt, MD

BRCA1 methylation or silencing does not predict for a better response to carboplatin over docetaxel in women with advanced triple-negative or BRCA1/2 breast cancer, according to results from a pre-planned biological analysis of the phase III TNT trial.

The results, presented at the 2016 San Antonio Breast Cancer Symposium, do not support testing for BRCA1 methylation or mRNA silencing to guide treatment in metastatic triple-negative breast cancer (TNBC), said Andrew Tutt, MD, a consulting clinical oncologist, Guy’s Hospital Breast Unit, and professor of oncology, King’s College London.

Previously, patients in the TNT trial with TNBC and germline BRCA1/2 mutations were found to have a higher response rate to carboplatin than docetaxel, and longer progression-free survival (PFS) rates with carboplatin. In the biologic analysis of the TNT trial, patients with tumors with BRCA1 methylation actually had a higher response to docetaxel than to carboplatin (42.1% vs. 21.4%, respectively; P = .28).

Patients enrolled in TNT had estrogen receptor-negative, progesterone-receptor-negative, HER2-negative breast cancer or had a known germlineBRCA1/2mutation. 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.

Thea priorisubgroup analyses were performed around “BRCAness” populations, said Tutt, including those with germlineBRCA1/2mutations and those with mutational signatures of homologous recombination (HR) deficiency, including patients with epigenetic dysregulation ofBRCA1through methylation and silencing of theBRCA1mRNA.

“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 ofBRCA1is known to be subject to such epigenetic silencing.”

Methylation in the regulatory region ofBRCA1occurs in about 10% to 40% of TNBCs. CpG methylation above the 10% level is associated with silencing ofBRCA1mRNA.

Thea priorihypothesis was thatBRCA1methylation would show a significant interaction with carboplatin treatment effect, as was done for the germlineBRCA1/2population.

DNA from 224 primary tumors was subjected toBRCA1methylation 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 beBRCA1methylated and 179 were non-methylated.BRCA1methylation and germlineBRCA1mutation co-existed in 2 patients.

“Counter to our hypothesis, those patients withBRCA1methylation had a higher response to docetaxel than to carboplatin,” said Tutt. The objective response rates (ORRs) were 21.4% in theBRCA1methylated patients randomized to carboplatin compared with 42.1% in those randomized to docetaxel. The difference did not achieve significance (P= .28).

In patients withoutBRCA1methylation, 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 byBRCA1methylation status did not achieve significance (P= .35).

BRCA1methylation had a different relationship with treatment response to germlineBRCAmutation. Response rates in the 31 patients with BRCA methylated and germlineBRCAwild type were 15.4% in the carboplatin group versus 44.4% in the docetaxel group (P= .13). In the 43 patients with germlineBRCAmutations, 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 forBRCA1mRNA silencing status, and results were available for 191. A cutpoint of value of log2(mRNA)<8.4 was established forBRCAmRNA 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 byBRCAmethylation, “we again showed, counter to our hypothesis, that theBRCA1silenced 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 theBRCA-mutated population did not change the overall result.

Unlike with germlineBRCA1/2 status, there was no significant effect ofBRCAmethylation on PFS following either carboplatin or docetaxel.

Exploratory analyses examining potential relationships between a response to carboplatin and new cutpoints forBRCA1methylation orBRCA1mRNA 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 forBRCA1/2have a high degree of homogeneity for mutation and functional loss through loss of the wild-type allele. The germline-mutated tumor retains itsBRCAmutation in the micrometastasis, and at relapse, retains its mutation and HR defect, rendering it more sensitive to platinum chemotherapy.

Reference:

Tutt A, Cheang MCU, Kilburn L, et al.BRCA1methylation 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.