Exploring the effects of exercise on immune cell function and tumour infiltration in patients with breast cancer receiving neoadjuvant chemotherapy - a feasibility trial

Abstract

Pre-clinical studies have shown that exercise can decrease tumour growth through mobilisation, activation, and increased tumour infiltration of natural killer (NK) and CD8⁺ T cells. It is currently unclear whether this can be extrapolated to patients. Therefore, a pilot study was set up to examine the feasibility of obtaining an additional study biopsy and to generate preliminary data on the potential effects of exercise on peripheral immune cell function and tumour immune infiltration. Twenty patients with stage I-III breast cancer receiving neoadjuvant chemotherapy were included (participation rate: 27%). Patients were randomised into the intervention group receiving a six-week supervised aerobic and resistance exercise program or the control group. Blood samples and tumour biopsies were collected before randomisation and after six weeks of chemotherapy. For 8 of 20 (40%) patients, we were able to obtain and analyse biopsies at diagnosis and six-week follow-up. This showed a decrease in CD56⁺ cells/mm² tumour tissue in the three patients of the control group, while it remained stable in most patients of the exercise group. Upon co-culture of peripheral blood mononuclear cells with K562 tumour cells, the exercise group showed increased expression of the degranulation marker CD107a on NK cells (β = 1038.5, 95%CI = 56.9; 2020.2, p = 0.04), and a trend towards increased tumour cell lysis in vitro (β = 18.8%, 95%CI = -3.9; 41.5, p = 0.10) compared to the control group. In conclusion, the study design was feasible with regard to the participation rate, however, revision is needed with regard to the use of a study-related biopsy prior to a sufficiently powered randomised controlled trial.

Keywords: Cancer; Chemotherapy; Exercise; Immune monitoring.

Fig. 1

Patient flowchart of the SPRINT pilot study.¹Reasons for missing fitness tests were related to low neutrophil values (n = 1) or refusal due to poor mental wellbeing (n = 1).²Three questionnaires were not returned upon request. ³Reasons for missing biopsies were pathological complete response (n = 2), decline due to fear and poor mental wellbeing (n = 2), too low neutrophil values for biopsy (n = 1), or disproportionate side effects from diagnostic biopsy (n = 1). ⁴One patient was excluded from the immune analyses due to co-medication with salmeterol, which has been shown to affect NK cell phenotype and function. ⁵Due to a low amount of isolated PBMCs, in one case only NK cell phenotype panel 1 could be measured and in two cases no NK cell phenotype panels could be measured. ⁶Three biopsies could not be analysed due to limited (<0.1 mm²) tumour tissue present in the biopsy. For one additional patient, the CD4/CD8 ratio could not be assessed due to too much necrotic tissue in the biopsy.

Fig. 2

Changes in immune profile from baseline to follow-up in the exercise intervention and control groups. A) Absolute counts of leukocytes; B) absolute counts of lymphocytes; C) %monocytes of peripheral blood mononuclear cells (PBMCs); D) %B cells of lymphocytes; E) %Natural Killer (NK) cells of lymphocytes; F) %CD56^dim of NK cells; G) %CD56^bright of NK cells; H) %NKT cells of lymphocytes; I) %T cells of lymphocytes; J) %regulatory T cells (Treg) of T cells; K) %CD4⁺ of T cells; L) %CD8⁺ of T cells; M) %T_naïve, %T_CM, %T_EM and %T_EMRA of CD4⁺ T cells; N) %T_naïve, %T_CM, %T_EM and %T_EMRA of CD8⁺ T cells; and O) %T_naïve of CD8⁺ cells in peripheral blood of patients with breast cancer from the exercise intervention (n = 7) and control group (n = 9) before (T0) and after (T1) six weeks of neoadjuvant chemotherapy. Data is presented as individual values and/or mean. Statistically significant differences (p < 0.05) are indicated by ∗. Abbreviations: ns, not significant; CM, central memory; EM, effector memory; EMRA, effector memory CD45RA⁺.

Fig. 3

Changes in peripheral NK cell phenotype from baseline to follow-up in the exercise intervention and control groups. The absolute change in percentage positive cells between baseline (T0) and after six weeks of neoadjuvant chemotherapy (T1) of NK cell markers on peripheral A) CD56^dim and B) CD56^bright NK cells of patients with breast cancer from the exercise intervention (n = 4 or 6) and control group (n = 9). A trend towards a lower expression of the immune checkpoint receptor CD96 (β = −8.2%, 95%CI = −16.8; 0.5, p = 0.06) and the activating receptor NKG2D (β = −3.7%, 95%CI = −7.7; 0.3, p = 0.06) on CD56^dim NK cells was observed in the exercise group compared to the control group after 6 weeks of neoadjuvant chemotherapy. Green = activating, orange = activating/inhibitory, red = inhibitory. Data is presented as individual values (ΔT, %positive NK cells at T1 – %positive NK cells at T0). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 4

Changes in peripheral NK cell function from baseline to follow-up in the exercise intervention and control groups. The percentage positive NK cells and median fluorescence intensity (MFI) of A, B) the degranulation marker CD107a; C, D) the early activation marker CD69; E, F) the cytotoxicity marker granzyme B; G, H) the cytokine IFN-γ; and I, J) the cytokine TNF after co-culture of unstimulated PBMCs of the exercise intervention (n = 7) and control group (n = 9) with K562 cells at a 10:1 ratio before (T0) and after (T1) six weeks of neoadjuvant chemotherapy. K) The ΔLysis (%) (T1 – T0) of K562 tumour cells after 4h co-culture with unstimulated PBMCs. %Lysis was calculated as the percentage killed K562 cells, corrected for spontaneous K562 cell death. Individual values and mean values are presented. Trends towards statistically significant differences (p < 0.10) are indicated by #, and statistically significant differences (p < 0.05) are indicated by ∗.

Fig. 5

Changes in immune cell infiltration in the tumour from baseline to follow-up in the exercise intervention and control groups. A) The number of CD56⁺ cells per mm² tumour tissue and B) the CD4/CD8 ratio in the biopsies of the exercise intervention (n = 5) and control group (n = 3 or n = 2) before (T0) and after (T1) six weeks of neoadjuvant chemotherapy. Trends towards statistically significant differences (p < 0.10) based on exploratory linear regression analyses are indicated by #, and statistically significant differences (p < 0.05) are indicated by ∗.