More T Cells versus Better T Cells in Patients with Breast Cancer

As in many other cancer types, the immune system modulates progression of breast carcinoma. However, immunotherapy has not yet achieved major breakthroughs, despite promising early-phase clinical trial results, particularly in patients with triple-negative breast cancer (in which the cancer cells are negative for estrogen and progesterone receptors and HER2; ref. 1). Failures of T cell–based therapies also occur in many patients with cancers from other organs, even in the field of skin melanoma that plays a leading role in innovative immunotherapies (2).

It is likely that cancer types that are resistant to today's immunotherapies will become responders to future treatments. This assumption is based on recent progress demonstrated with drug combinations that promote immune responses and simultaneously block cancer cell aggression, stromal and extracellular matrix condensing, vessel distortion, and/or immune suppression. Therefore, further research is necessary to identify mechanisms that are of key pathogenic importance in specific patients and to modulate them by specific therapies.

A central role in immune protection against cancer is played by T lymphocytes, particularly CD8⁺ T cells (3). Their infiltration in tumor cell nests is usually associated with a favorable prognosis and may predict outcome of therapies with drugs that block immune inhibitory receptors (checkpoint blockade; refs. 4, 5). A recent meta-analysis reported that tumor-infiltrating CD8⁺ lymphocytes (TIL) can be identified in 48% of all breast cancers (6). Interestingly, triple-negative breast cancers show the highest incidence of lymphocyte predominance. Moreover, TILs were found to be prognostic in triple-negative breast cancer, and higher levels of TILs were predictive for trastuzumab benefit in HER2⁺ disease (7). These findings suggest that therapies that enforce immune responses could potentially improve patient survival. But why are the results of immunotherapy trials not better?

A new study by Gil Del Alcazar and colleagues in this issue of Cancer Discovery takes a closer look at the immune landscape of breast cancer (8). It describes not only fully developed invasive ductal carcinomas (IDC) but also normal breast tissues and ductal carcinoma in situ (DCIS) that have attracted limited attention of immunologists in the past. T cells were found to be the most dominant hematopoietic (CD45⁺) cell population in both normal and cancerous breast tissues. The authors report that in situ carcinomas were enriched for T-cell gene sets, including high diversity of T-cell receptor (TCR) clonotypes, more activated CD8⁺ T cells, and more T cells expressing the inhibitory receptor TIGIT. Interestingly, T cells were rarely in direct contact with tumor cells at this early stage of the disease. In contrast, fully established IDCs were highly infiltrated by T cells, were enriched for regulatory T-cell gene sets, and had fewer activated CD8⁺ T cells and less diverse TCR clonotypes, but higher expression of the immune inhibitory ligand PD-L1.

The authors show that leukocyte infiltration (“inflammation”) increased with tumor progression in breast cancer, as previously described for other tumors. T cells were more numerous in IDC as compared with early stages, illustrating that enhanced lymphocyte infiltration does not necessarily reflect stronger antitumor immune responses. It is generally known that higher percentages of TILs are associated with increased immune-suppressive functions by various cell populations such as stromal cells, myeloid cells, or regulatory T cells. Indeed, a different picture emerged when the authors focused on functional T-cell features instead of only counting their numbers. In this regard, the correlations were inverse, providing evidence for better T-cell function in early disease stage (DCIS), with significantly higher fractions of CD8⁺ T cells expressing Ki67, Granzyme B, and IFNγ, associated with the T cell's potential to proliferate, kill, and express effector cytokines, respectively (Fig. 1). As usual, and despite using high-end laboratory technologies, such correlative studies do not prove cause–consequence relationships, leaving open the question of whether declining functional T-cell competence is co-responsible for disease evolution, or whether the observed T-cell properties are solely the result of the pathophysiologic conditions of the different disease types and stages, without directly affecting disease progression.

The authors also studied genomic alterations of cancer cells. In some patients, they found coamplification of the 17q12 chemokine cluster region with ERBB2 in HER2⁺ breast tumors, negatively correlating with activated intratumoral T cells. Special emphasis was given to the increasingly recognized 9p24 amplicon that includes JAK2 and CD274/PDCD1LG2 (encoding PD-L1/PD-L2). The authors found CD274 copy-number gain and PD-L1 overexpression in 3 of 10 patients with triple-negative IDC (but in none of the patients with DCIS), suggesting selection of these tumor cells due to immune resistance. This discovery is of possible relevance for therapy choice for triple-negative breast cancer, according to findings in Hodgkin lymphoma, where this same 9p24 amplicon is strongly predictive of clinical response to anti–PD-1 therapy.

For a better understanding of immune (inhibitory) mechanisms in patients with cancer, it will be generally useful to determine whether susceptibility versus resistance to immunotherapy is caused by genomic or epigenetic alterations of tumor cells and outgrow of selected variants. Alternatively, and perhaps more frequently, immune inhibition is due to conserved (natural) regulation of T-cell responses. Upregulation of the immune inhibitory ligand PD-L1 is a typical example, as it is quite systematically upregulated by IFNγ secreted by intratumoral lymphocytes. Consequently, several reasons can account for high PD-L1 expression: either genomic alterations, conserved pathways, or a combination of multiple factors (3).

The new study (8) represents a significant advancement in the characterization of immune parameters in human breast cancer. It is also the first comprehensive study in a human preinvasive lesion that includes flow cytometry and gene expression profiling with follow-up in situ validation and analysis of the cancer cell's genetic changes. Although the majority of findings need further refinements and ultimately elucidation of the underlying mechanisms, the results provide specific guidance for identifying causal relationships and biomarkers.

Ultimately, the demonstration of disease-driving mechanisms requires biological analyses of pretreatment versus posttreatment specimens revealing cellular and molecular changes that cause therapy resistance. It is important to tackle the practical challenges of obtaining and processing multiple and sequential biopsy specimens per patient. Of course, basic research in animals can and must also substantially guide investigations. However, models that reliably reproduce human disease are not (yet) available, for example, for DCIS.

Increasing the analytic resolution for identifying cell subsets will support progress. Given the usually small numbers of cells that can be obtained from tumor specimens, the novel molecular single-cell techniques are promising (9). Different subsets of CD8⁺ T cells (naïve, effector, memory, activated, exhausted) and CD4 helper T cells (Th1, Th2, and Th17 cells, and regulatory T cells) play in part fundamentally different roles that remain largely unclear in breast cancer, similar to the majority of other cancers. Molecular single-cell studies are beginning to reveal the great complexity of cellular subsets and their molecular and functional properties. Single-cell data can ideally be complemented with multiplexed histologic analyses (10), together providing microanatomic insights, information on potential functions, and possible interactions between the cells of the tumor microenvironment. The great puzzles of immune pathophysiologic mechanisms in patients with cancer will be elucidated step by step, providing an increasingly sound rationale to design clinical trials.

The search is on for new biomarkers pinpointing patients who benefit significantly from tailored immunotherapy. The experience with targeted therapies (with, e.g., oncogene blockers) teaches an important lesson, namely that therapy choices are now well guided by the molecular settings of individual tumors, validated in clinical trials. Similar principles will be implemented for the development of immunotherapies, as they can be optimized based on results from prospective trials that determine the importance of immune biological features for individual patient care.

No potential conflicts of interest were disclosed.

This work was supported in part by the University of Lausanne, Ludwig Cancer Research, the Cancer Research Institute, the Wilhelm Sander Foundation, the Swiss Cancer League (3679-08-2015 and 3971-08-2016), the Swiss National Science Foundation (CRSII3_160708 and 320030_152856), Alfred and Annemarie von Sick, and the Max Cloëtta Foundation.