Intrinsic Tumor Subtyping Is Key To Breast Cancer Treatment

Charles M. Perou, PhD, discusses the role of intrinsic tumor subtyping in treatment decisions and outcomes for patients with breast cancer.

The importance of identifying the genomic intrinsic subtype of breast cancer tumors cannot be overstated, Charles M. Perou, PhD, said during the 21st Annual International Congress on the Future of Breast Cancer® (IBC) East. Intrinsic subtype can predict treatment response and prognosis, particularly for basal-like subtype tumors within non–triple-negative breast cancers (TNBCs).

“I will continue to make the argument that knowing a tumor subtype is important, particularly when you find a basal-like cancer in an area where you don't think you're going to find it,” Perou said. “Within hormone receptor [HR]­–positive, HER2-negative [disease there is] potential benefit of CDK4/6 inhibitors varying by subtype. For [luminal] A and B, there is a benefit and for basal-like, there’s not a benefit. In HER2-enriched [disease], the kind of CDK4/6 [inhibitor] may matter. We’ll see.”

Perou is the May Goldman Shaw Distinguished Professor of Molecular Oncology, codirector of the Computational Medicine Program, faculty director of the UNC Lineberger Bioinformatics Group, codirector of the UNC Lineberger Breast Cancer Research Program, and a professor in the department of genetics at the University of North Carolina Lineburger Comprehensive Cancer Center. He is also the 2019 Giants of Cancer Care® award winner for cancer diagnostics and delivered the Giants lecture at IBC East.

Perou said the luminal A and B subtypes appear in up to 70% of breast cancers, including most estrogen receptor– and progesterone receptor –positive tumors. These subtypes are prognostic for outcomes and associated with a treatment benefit from endocrine therapy and chemotherapy.

The basal-like subtype is present in 10%-15% of tumors, but it is seen in up to 75% of TNBCs. The overwhelming majority of TNBC are P53 deficient. They are typically highly proliferative and tend to be immune infiltrated.

“When we’re studying TNBC, we’re really studying basal-like breast cancers,” Perou said. “If we’re going to make an impact against the majority of TNBCs, we really have to study and attack the biology of basal-like breast cancers, which includes an association with BRCA1 mutation status.”

Finally, there is the HER2-enriched subtype, which is present in 10%-15% of tumors. This subtype is defined in part by amplification of HER2 region on 17q11-12. More than 70% of tumors in this subtype of TP53­-mutated, and this subtype is predictive of benefit for HER2-targeting agents and combinations. “Almost all HER2-enriched tumors, but not all, are HER2 amplified and as you'll see, certainly not all HER2-amplified tumors fall into this expression subtype.”

Conventional wisdom says TNBCs highly heterogeneous. However, Perou argued that TNBCs are actually the most homogeneous group, while HER2-positive tumors are the most heterogenous. “They have all the subtypes present, and all of them present that a fairly decent frequency. And that’s going to matter,” he said.

He cited data from the phase 3 MONALEESA-2 (NCT01958021), MONALEESA-3 (NCT02422615), and MONALEESA-7 (NCT02278120) trials to illustrate the prognostic relationship of PAM50-based subtypes with progression-free survival (PFS) and risk of disease progression by subtype and treatment. Patients were treated with endocrine therapy and ribociclib (Kisqali)

for HR-positive, HER2-negative advanced breast cancer.1

Subtype distribution was 46.7% for luminal A (LumA), 24.0% for luminal B (LumB), 14.0% for normal like, 12.7% for HER2-enriched, and 2.6% for basal like. Investigators profiled 1160 tumors, with 672 patients assigned to ribociclib and 488 assigned to placebo. Distribution of subtypes was generally consistent across treatment arms and trials.

Prat et al found that associations between subtypes and PFS were statistically significant in both arms (P < .001). All subtypes except basal-like (HR, 1.15; P = .77) derived significant PFS benefit with ribociclib: HER2-enhanced (HR 0.39; P < .0001), LumB (HR, 0.52; P < .0001), LumA (HR, 0.63; P = .0007), and normal-like (HR, 0.47; P = .0005).

“Typically, basal-like in the HER2-enriched [subtype], particularly HER2-enriched in the absence of HER2 targeting, is the worst player,” Perou said. “Now, the interesting part of this study was when [they] looked at the benefit, these are randomized trials with and without the CDK4/6 inhibitor, and in this case, you can see there’s a benefit of the CDK4/6 inhibitor in [the] luminal A, [luminal] B, HER2-enriched [subtypes], but not in basal-like.”

Similarly, findings from the phase 3 CALGB 40601 trial (NCT00770809) showed that dual HER2-targeting agents induced a benefit for relapse-free survival (RFS) and pathologic complete response (pCR) compared with single targeting. Investigators assessed trastuzumab (Herceptin) plus lapatinib (Tykerb; n = 118), trastuzumab alone (n = 120), or lapatinib alone (n = 67) in women with untreated stage II and III HER2-positive breast cancer.2

At a median follow-up of 83 months, patients in the combination arm had significantly better RFS compared with trastuzumab monotherapy (HR, 0.32; 95% CI, 0.14-0.71; P = .005). There was no difference between the monotherapy arms.

Seven-year OS rates were 84% for lapatinib, 88% for trastuzumab, and 96% for the combination. OS was significantly higher in the combination arm compared with the trastuzumab arm (HR, 0.34; 95% CI, 0.12-0.94; P = .037).

Eighty-nine (61%) of 146 HER2-enriched patients had pCR in the breast compared with 29 (25%) of 118 non–HER2-enriched patients (odds ratio, 3.8; 95% CI, 2.23-6.72; P < .001). Investigators identified significant RFS differences were found among the different subtypes. Luminal A tumors had lowest pCR rate (14.3%) but carried the best RFS outcome, with no events recorded after 7 years of follow-up. In contrast, HER2-enriched patient had the highest pCR rate but a significantly worse RFS outcome, with 30 (20.5%) of 146 RFS events recorded.

Conflicting Findings in TNBC

In long-term results from the phase 2 CALGB 40603 trial (NCT00861705), a 2 × 2 randomized study, investigators determined that adding carboplatin and/or bevacizumab (Avastin) to standard neoadjuvant chemotherapy improved pCR rates compared with chemotherapy alone in TNBC.3 However, data published in 2019 and 2022 did not show a subsequent benefit for event-free survival (EFS).4,5

Perou said that finding was both disappointing and in contrast with results from the phase 3 BrighTness trial (NCT02488967). Those data showed that carboplatin plus veliparib (ABT-888) with paclitaxel improved both pCR and EFS compared with paclitaxel alone (HR, 0.63; 95% CI, 0.43-0.92, P = 0.02) at a median follow-up of 4.5 years.6

“Time permitting, I could spend 30 minutes discussing possible hypotheses about these, which was different than what we saw [in] BrighTness,” he said. “We’re actually in the midst of doing a combined analysis of BrighTness and [CALGB 40603] with genomics and clinical data, and so hopefully, we can learn something about that discrepant result between those 2 studies.”

Investigators analyzed more than 850 clinical and genomic features for association with outcomes in CALGB 40603. Only 27 were associated with both. Twenty-four of the features associated with both pCR and EFS reflected the tumor's immune microenvironment including the presence of a variety of immune effector cells, such as T and B lymphocytes and natural killer cells. Investigators also determined that higher mRNA expression levels of immune checkpoint genes, including tumor infiltrating lymphocytes (TILs), PD-1, and PD-L1 were associated with improvements in both pCR and EFS.

However, Perou noted that investigators performed multivariate analysis, adding TILs, age, pCR status, and stage to the base model to predict EFS. TILs were not a predictor for EFS, but the addition CD8 T-cell signature was.

“I believe that the gene expression signatures of the immune features are better predictors of event free survival than our than our TILs,” Perou said. “I’m going to claim also they’re more objective and more reproducible. Amongst triple negatives, there’s Pam50 subtypes, there’s the Vanderbilt subtypes, there’s the MD Anderson subtypes. Long story short, I'm going to argue they didn’t really matter.”


  1. Prat A, Chaudhury A, Solovieff N, et al. Correlative biomarker analysis of intrinsic subtypes and efficacy across the MONALEESA phase III studies. J Clin Oncol. 2021;39(13):1458-1467. doi: 10.1200/JCO.20.02977
  2. Fernandez-Martinez A, Krop IE, et al. Survival, pathologic response, and genomics in CALGB 40601 (Alliance), a neoadjuvant phase III trial of paclitaxel-trastuzumab with or without lapatinib in HER2-positive breast cancer. J Clin Oncol. 2020;38(35):4184-4193. doi: 10.1200/JCO.20.01276
  3. Sikov WM, Berry DA, Perou CM, et al. Impact of the addition of carboplatin and/or bevacizumab to neoadjuvant once-per-week paclitaxel followed by dose-dense doxorubicin and cyclophosphamide on pathologic complete response rates in stage II to III triple-negative breast cancer: CALGB 40603 (Alliance). J Clin Oncol. 2015;33(1):13-21. doi: 10.1200/JCO.2014.57.0572
  4. Sikov WM, Polley MY, Twohy E, et al. CALGB (Alliance) 40603: Long-term outcomes (LTOs) after neoadjuvant chemotherapy (NACT) +/- carboplatin (Cb) and bevacizumab (Bev) in triple-negative breast cancer (TNBC). J Clin Oncol. 2019; 37, (suppl; abstr 591). doi: 10.1200/JCO.2019.37.15_suppl.591
  5. Shepherd JH, Ballman K, Polley MC, et al. CALGB 40603 (Alliance): Long-term outcomes and genomic correlates of response and survival after neoadjuvant chemotherapy with or without carboplatin and bevacizumab in triple-negative breast cancer. J Clin Oncol. 2022;40(12):1323-1334. doi: 10.1200/JCO.21.01506
  6. Geyer CE, Sikov WM, Huober J, et al. Long-term efficacy and safety of addition of carboplatin with or without veliparib to standard neoadjuvant chemotherapy in triple-negative breast cancer: 4-year follow-up data from BrighTNess, a randomized phase III trial. Ann Oncol. 2022;33(4):384-394. doi: 10.1016/j.annonc.2022.01.009