Adjuvant T-DM1 Shows Consistent Benefit in Residual Invasive HER2+ Breast Cancer, Irrespective of Mutational Status

Ado-trastuzumab emtansine (T-DM1; Kadcyla) demonstrated a consistent invasive disease-free survival (iDFS) benefit versus trastuzumab (Herceptin) in patients with HER2-positive breast cancer across all biomarker subgroups evaluated in the pivotal phase 3 KATHERINE trial (NCT01772472), according to data from exploratory analyses presented during the 2020 ASCO Virtual Scientific Program.¹

“These exploratory analyses provide the first data on the relationship between biomarker expression and iDFS in residual disease after HER2-targeted therapy,” Carsten Denkert, MD, a professor at Universitätsmedizin Berlin Institute of Pathology, said in a presentation during the meeting.

Results showed that when looking at 4 different gene expression signatures—HER2, PD-L1, CD8, and T-cell effector—a consistent benefit with T-DM1 was observed, irrespective of biomarker subgroup. No difference was observed between the 2 treatment arms with regard to the expression of those 4 markers.

With regard to combined mRNA signatures of genes that play a key role in the cancer immunity cycle and in the interaction of tumor cells and immune cells, no differences were observed between the T-DM1 and trastuzumab arms. “The iDFS benefit was observed regardless of the status of those immune genes,” noted Denkert.

Previously, in May 2019, the FDA approved T-DM1 for use as an adjuvant treatment in patients with HER2-positive early breast cancer with residual invasive disease following neoadjuvant trastuzumab and chemotherapy based on earlier findings from KATHERINE which showed a 50% reduction in the risk of invasive disease recurrence or death versus trastuzumab in this setting (hazard ratio [HR], 0.50; 95% CI, 0.39-0.64; P <.0001).² The 3-year iDFS rate with T-DM1 versus trastuzumab was 88.3% and 77.0%, respectively; this translated to an absolute improvement of 11.3%.

In the open-label KATHERINE trial, a total of 1,486 patients with centrally confirmed HER2-positive, nonmetastatic, invasive primary breast cancer who had residual invasive tumor in the breast or axillary nodes at surgery following completion of a minimum of 6 cycles of neoadjuvant chemotherapy comprised of trastuzumab plus a taxane with or without an anthracycline and then surgery.

Following surgery, subjects were then randomized 1:1 to receive 14 cycles of either trastuzumab (n = 743) or T-DM1 (n = 743). Tissue samples were collected from patients with a focus on post-neoadjuvant surgery samples; 80% of patients had such surgical samples available. For the remaining 20% of patients, pre-therapeutic samples were collected, and were utilized for some of the biomarker analysis, according to Denkert.

The translational program was comprised of 2 parts. In the first part, investigators analyzed PIK3CAmutations through DNA sequencing; a total of 1,363 samples were available for evaluation. In the second portion of the program, the focus of study was mRNA expression through RNA sequencing; 1,059 samples passed quality control checks based on library size and mapping quality. Of those samples, 244 were pre-neoadjuvant samples and 815 were post-neoadjuvant surgical samples.

Investigators found that the post-neoadjuvant samples were representative of the entire intent-to-treat (ITT) patient population, and as such, the statistical and biomarker analysis for markers such as HER2, PD-L1, CD8, and predefined immune signatures, were performed by using those post-neoadjuvant surgical samples. The analysis was adjusted for tumor content. The multivariate analysis was adjusted for tumor content, clinical stage, hormone receptor status, previous HER2-directed therapy, and pathologic nodal status.

“In the translational program, we focused on those pathways that are known to be relevant for resistance and response to HER2 treatment. On the one hand, there was a HER2 pathway. We know that higher HER2 expression is associated with improved outcome after HER2-targeted therapy,” explained Denkert. “However, the data are based on pre-therapeutic samples; we do not know whether the same is still true if we look at the residual resistant post-neoadjuvant tumor. On the other hand, we know that PIK3CA is an important element of cellular pathways and is a resistance factor and known marker of poor prognosis in metastatic breast cancer.”

When investigators looked at PIK3CA mutations, no notable difference in treatment between mutated and non-mutated tumors was observed. In the ITT population (n = 743), the 3-year iDFS was 88.3% in those who received T-DM1 versus 77.0% in those who were given trastuzumab (HR, 0.50; 95% CI, 0.39-0.64). For those with mutated tumors who received T-DM1 and trastuzumab, the iDFS rates were 88.9% versus 77.9%, respectively (HR, 0.54; 95% CI, 0.32-0.90); in those with non-mutated tumors the rates were 88.3% versus 77.0%, respectively (HR, 0.48; 95% CI, 0.35-0.65).

“The 3-year iDFS rates were nearly identical between the mutated and non-mutated tumors,” noted Denkert. “In addition, we did not observe a prognostic effect of PIK3CA mutations in this cohort.”

Next, investigators looked at HER2 mRNA to compare the gene expression between pre- and post-neoadjuvant surgical samples. Here, the post-neoadjuvant surgical samples showed a decreased expression of HER2 mRNA versus the pre-neoadjuvant samples (P <.0001). “However, there was this interesting population where there was still a high level of HER2 mRNA in the [post-neoadjuvant] surgical, and therefore, therapy-resistant, samples,” said Denkert.

When looking at PD-L1 and other immune marker gene expression, no differences were observed between pre-neoadjuvant and post-neoadjuvant surgical samples.

Data for the baseline characteristics between the surgical biomarker population and the ITT biomarker population, showed no differences for treatment arm, clinical stage at presentation, hormone receptor status, type of HER2 therapy, and pathological nodes, noted Denkert. In the ITT population (n = 1486), the HR with T-DM1 was 0.50 (95% CI, 0.39-0.64) versus 0.39 in the surgical biomarker population (n = 815; 95% CI, 0.28-0.55).

“The benefit of T-DM1 was very similar in the ITT population and in the surgical biomarker population, so this is a representative population for biomarker of analysis,” Denkert said.

When investigators looked more closely at HER2 gene expression in the post-neoadjuvant samples, they noted a difference between the treatment arms. Patients who were treated with trastuzumab and who had a tumor with high HER2 expression in the post-neoadjuvant residual tumor experienced a worse outcome with worse iDFS. “This effect was not seen in the T-DM1 arm,” noted Denkert.

With regard to HER2 gene expression, the HRs for the high-versus-low effect for T-DM1 and trastuzumab were 1.01 (95% CI, 0.56-1.83) and 2.02 (95% CI, 1.32-3.11), respectively. Notably, the biomarker impact in the trastuzumab arm continued to be significant after adjusting for other factors in the multivariate analysis.

“What we see here is a kind of resistant population; residual tumors that have a high HER2 expression in this setting are resistant to trastuzumab, but not to T-DM1,” said Denkert.

Then, investigators looked at the remaining immune markers and gene signatures. With regard to PD-L1 expression, they found that low PD-L1 expression in residual tumors was associated with a worse outcome for those who took trastuzumab; however, those in the T-DM1 arm did not experience the same outcome. “This may indicate that PD-L1 may be involved in some resistance mechanisms,” concluded Denkert.

References

1. Denkert C, Lambertini C, Fasching PA, et al. Biomarker data from KATHERINE: a phase III study of adjuvant trastuzumab emtansine (T-DM1) versus trastuzumab (H) in patients with residual invasive disease after neoadjuvant therapy for HER2-positive breast cancer. J Clin Oncol. 2020;38(suppl 15):502. doi: 10.1200/JCO.2020.38.15_suppl.502
2. Von Minckwitz G, Huang C-S, Mano MS, et al. Trastuzumab emtansine for residual invasive HER2-positive breast cancer. N Engl J Med. 2019;380(7):617-628. doi:10.1056/NEJMoa1814017