Enhanced humoral immunity in breast cancer patients with high serum concentration of anti-HER2 autoantibody

Abstract

Humoral immunity plays a substantial role in the suppression of breast cancer. We have revealed that a high serum concentration of anti-HER2 autoantibody (HER2-AAb) is associated with favorable outcomes in patients with invasive breast cancer. Thus, we aimed to clarify the association between high serum concentration of HER2-AAb and humoral immune response in the tumor microenvironment. Out of 500 consecutive patients with invasive breast cancer, we selected those whose HER2-AAb values were high (n = 33) or low (n = 20) based on the distribution of HER2-AAb values of 100 healthy individuals. Tumor and regional lymph node formalin-fixed paraffin-embedded samples prepared from the surgical specimens were subjected to immunohistochemistry. We confirmed that the recurrence-free survival of the high HER2-AAb group was significantly longer than that of the low HER2-AAb group (p = 0.015). The numbers of tumor-infiltrating CD20+ immune cells (ICs) (p < 0.001), IGKC+ICs (p = 0.023), and CXCL13+ ICs (p = 0.044) were significantly higher in the high HER2-AAb group than in the low HER2-AAb group. The number of CD4+ ICs in the B-cell follicles of the regional lymph nodes was also significantly greater in the high HER2-AAb group than in the low HER2-AAb group (p = 0.026). Our findings indicate that a high level of HER2-AAb is associated with enhanced humoral immunity against breast cancer and thus may provide a rationale for the association of HER2-AAb with favorable prognosis.

Keywords: B-lymphocytes; CD4-positive T-lymphocytes; humoral immunity; tumor microenvironment; tumor-infiltrating lymphocytes.

FIGURE 1

(A) Distribution of the log‐transformed serum anti‐HER2 autoantibody (HER2‐AAb) concentrations of 100 healthy individuals. Shapiro‐Wilk test was performed to demonstrate the normality of the distribution. (B) An example of IC counting. FOXP3‐positive ICs infiltrating the IT (black arrowheads), AS (white arrowheads), and distant stromal area (red arrowheads) are indicated. The area surrounded by a broken line denotes the tumor stroma. Original magnification, ×40. IC, immune cell; IT, intratumor area; AS, adjacent stromal area. (C) Kaplan–Meier curves of patients with high HER2‐AAb concentration (red line, HER2‐AAb >67.81 ng/ml), intermediate HER2‐AAb concentration (green line, 2.27 ng/ml ≤HER2‐AAb ≤67.81 ng/ml), and low HER2‐AAb concentration (blue line, HER2‐AAb <2.27 ng/ml). Percentages of recurrence‐free survival are shown. p, log‐rank test. Scale bar, 100 µm

FIGURE 2

Tumor‐infiltrating lymphocytes (TILs), PD‐L1‐positive ICs, CD8‐positive ICs, FOXP3‐positive ICs, and CD8‐positive‐to‐FOXP3‐positive ratios in the high and low HER2‐AAb groups. (A and B) Representative photographs of hematoxylin‐eosin staining of the tumor with high TIL content (A) and of the tumor with low TIL content (B). (C–H) Representative photographs of immunohistochemical (IHC) staining for PD‐L1 (C and D), CD8 (E and F), and FOXP3 (G and H) of the tumors with large cell counts (C, E and G) and the tumors with small cell counts (D, F and H). Original magnification, ×20. (I) Percent TIL content in the high HER2‐AAb group and the low HER2‐AAb group. (J–L) Counts of ICs positive for PD‐L1 (J), CD8 (K), and FOXP3 (L) in the high HER2‐AAb group and the low HER2‐AAb group. Total denotes the cells in IT, AS, and the distant stroma, while IT+AS denotes the cells in IT and AS. (M) CD8‐positive‐to‐FOXP3‐positive ratios in the high HER2‐AAb group and the low HER2‐AAb group. Medians are shown by horizontal dotted lines. p, Mann–Whitney U test. Scale bar, 100 µm

FIGURE 3

Tumor‐infiltrating CD20‐positive ICs, IGKC‐positive ICs, and CXCL13‐positive ICs, and follicular CD4‐positive ICs in the regional axillary lymph nodes in the high HER2‐AAb group and the low HER2‐AAb group. (A–F) Representative photographs of IHC staining for CD20 (A and B), IGKC (C and D), and CXCL13 (E and F) of the tumors with large cell counts (A, C and E) and the tumors with small cell counts (B, D and F). (G–J) Representative photographs of IHC staining for CD4 (brown) and CXCR5 (red) of the regional axillary lymph nodes with large cell counts (G and I) and with small cell counts (H and J). Images (I) and (J) are the high power fields of the images (G) and (H), respectively. Original magnification, ×20 (A–H) and ×40 (I and J). (K–M) Counts of ICs positive for CD20 (K), IGKC (L), and CXCL13 (M) in the high HER2‐AAb group and the low HER2‐AAb group. (N) Counts of CD4‐positive ICs in CXCR5‐positive B‐cell follicles (follicular CD4⁺) of the regional axillary lymph nodes in the high HER2‐AAb group and the low HER2‐AAb group. Medians are shown by horizontal dotted lines. p, Mann–Whitney U test. Scale bar, 100 µm (A–H) and 50 µm (I and J)

FIGURE 4

Repertoire analysis for B‐cell receptors (BCRs) in TILs. (A and B) Examples of IgHVJ repertoire 3D histograms (% usage). X‐axis and Y‐axis denote IgHV repertoire and IgHJ repertoire, respectively. The diversity of BCRs was estimated by Shannon entropy H. (C) Shannon entropies of BCR repertoires of tumor‐infiltrating ICs. Medians are shown by horizontal dotted lines

FIGURE 5

Tumor factors associated with HER2‐AAb. (A–F) Representative photographs of IHC staining for CXCL13 (A and B), HLA‐A/B/C (C and D), and PD‐L1 (E and F). Highly expressing tumor cells (A, C and E) and minimally expressing tumor cells (B, D and F) are shown. Original magnification, ×20. (G–I) Percent tumor cells positive for CXCL13 (G), HLA‐A/B/C (H), and PD‐L1 (I) in the high HER2‐AAb group and low HER2‐AAb group. Medians are shown by horizontal dotted lines. p, Mann–Whitney U test. Scale bar, 100 µm

FIGURE 6

(A–C) HER2 status in the high HER2‐AAb group and low HER2‐AAb group. (A) HER2 gene amplification estimated by fluorescence in situ hybridization (FISH). HER2 copy number was standardized with CEP17 and is demonstrated by the HER2‐to‐CEP17 FISH ratio. (B) HER2 expression estimated by the standard IHC scoring. (C) Missense somatic mutations in the HER2 gene. In 38 analyzed tumor samples, five tumors had missense mutations in the HER2 gene. ECD, extracellular domain; TM, transmembrane domain; TK, tyrosine kinase domain. (D and E) Association between two factors among all factors evaluated in the present study. Spearman's rank correlation coefficients r_S (D) and the corresponding p values (E) are shown by heat maps