The prognostic significance of peritumoral tertiary lymphoid structures in breast cancer

Abstract

Tumors and their surrounding area represent spatially organized "ecosystems", where tumor cells and the immune contextures of the different compartments are in a dynamic interplay, with potential clinical impact. Here, we aimed to investigate the prognostic significance of peritumoral tertiary lymphoid structures (TLS) either alone or jointly with the intratumoral densities and spatial distribution of CD8 + and CD163 + cells in breast cancer (BCa) patients. TLS were identified peritumorally, within the area distancing up to 5 mm from the infiltrative tumor border, counted and further characterized as adjacent or distal, in formalin-fixed, paraffin-embedded tumor tissue samples from a cohort of 167 patients, with histologically confirmed invasive ductal BCa. TLS and tumor-infiltrating immune cells were determined by H&E and immunohistochemistry. Clinical follow-up was available for 112 of these patients. Patients with peritumoral TLS exhibited worse disease-free survival (DFS) and overall survival (OS) as compared to patients lacking TLS. Moreover, the density of peritumoral TLS was found to be crucial for prognosis, since patients with abundant TLS exhibited the worst DFS and OS. By combining the density of adjacent TLS (aTLS) with our recently published intratumoral signatures based on the differential distribution of CD8 + and CD163 + in the tumor center and invasive margin, we created two improved immune signatures with superior prognostic strength and higher patient population coverage. Our observations strengthen the notion for the fundamental role of the dynamic interplay between the immune cells within the tumor microenvironment (center/invasive margin) and the tumor surrounding area (peritumoral TLS) on the clinical outcome of BCa patients.

Keywords: Breast cancer; Immune contexture; Prognostic signatures; Tertiary lymphoid structure; Tumor infiltration.

Fig. 1

Adjacent and distal TLS and their relation to clinicopathological characteristics. a Two aTLS and one dTLS, stained with anti-CD3 (× 40 magnification), are shown. The invasive margin (IM) of the tumor is also visible. b Shown is an aTLS next to the tumor with a cluster of cancer cells (cytokeratin staining, × 100 magnification). c Distribution of aTLS and dTLS-positive or -negative tumors among BCa patients stratified by molecular subtypes. Percentages of d aTLS and e dTLS-positive (+) or -negative (−) tumors in the total BCa patient population and in groups of patients stratified by clinicopathological characteristics. f Presence of aTLS within dTLS − (dTLS0) and dTLS + patients. (p < 0.05*, p < 0.01**, p ≤ 0.0001****)

Fig. 2

Adjacent and distal TLS and their relation to clinical outcome. Clinical outcome (Kaplan–Meier disease-specific survival curves) in patients with a aTLS − and aTLS + tumors or b dTLS − and dTLS + tumors. c Clinical outcome in patients with TLS-negative tumors (aTLS − dTLS −), or with tumors harboring only aTLS (aTLS + dTLS −) or both aTLS and dTLS aTLS (aTLS + dTLS +). d Statistical analyses among the indicated groups

Fig. 3

Number of aTLS and their relation to clinicopathological characteristics and clinical outcome of patients. a Clinical outcome in patients with tumors without aTLS (aTLS0) or with up to 4 aTLS (aTLS1–4) and ≥ 5 aTLS (aTLS ≥ 5). b Statistical analyses among groups. c Distribution of aTLS-negative (aTLS0) or aTLS-positive (aTLS1–4 and aTLS ≥ 5) tumors among BCa patients stratified by molecular subtypes. d Distribution of aTLS-negative (aTLS0) or aTLS-positive (aTLS1–4 and aTLS ≥ 5) tumors among BCa patients stratified by clinicopathological characteristics. (p < 0.05*, p < 0.01**)

Fig. 4

aTLS and dTLS distribution in the FCIS, UCIS, and REST patients groups. a Percentages of patients in the FCIS, REST, and UCIS groups with aTLS-negative (aTLS0) or aTLS-positive (aTLS1–4 and aTLS ≥ 5) tumors and b with dTLS-negative (dTLS0) or dTLS-positive (dTLS +) tumors. c Subgroups in the REST group with aTLS-negative (aTLS0) or aTLS-positive (aTLS1–4 and aTLS ≥ 5) tumors stratified by immune infiltrates in the TC and IM; LL/LL: low number of CD8 + in the TC and IM and low number of CD163 + in the TC and IM; HH/LL: high number of CD8 + in the TC and IM and low number of CD163 + in the TC and IM; HH/HH: high number of CD8 + in the TC and IM and high number of CD163 + in the TC and IM; LL/HH: low number of CD8 + in the TC and IM and high number of CD163 + in the TC and IM. d, e Clinical outcome in patients belonging to the REST group with tumors having aTLS0, aTLS1–4, or TLS ≥ 5. f Statistical analyses among groups

Fig. 5

The reinforced immune signatures. a Clinical outcome in BCa patients with RFCIS or RUCIS reinforced signatures. b %Probability and number of subjects at risk at 5 years for DFS and OS. c Distribution of the reinforced immune signatures in the total patient population and in patients’ groups based on clinicopathological parameters. d Distribution of the reinforced signatures among BCa patients with different molecular subtypes. e Densities of CD4, CD8, CD163, and FoxP3 cells within TLS in BCa patients having RFCIS or RUCIS. Statistical significance refers to comparison of the percentage of patients. (p < 0.05*, p < 0.01**)