Introduction
Oesophageal cancer is associated with unfavourable prognosis. Surgical resection is the best option when curation is aimed for. However, even after surgery with curative intent, overall survival is only 20–40% [
9,
26]. Many institutes apply neoadjuvant therapy to improve the long-term outcome, especially after the publication of favourable long-term results of a randomised trial from the Medical Research Council Oesophageal Cancer Working Party, comparing neoadjuvant chemotherapy followed by surgery versus surgery alone [
23]. However, in a large proportion of patients insufficient objective response is achieved. To improve these disappointing response rates, different protocols of preoperative chemoradiation therapy have been proposed and evaluated [
17]. Promising results have been achieved with the addition of heat to chemotherapy or radiotherapy. Hyperthermia is a treatment modality in which oesophageal tissue is exposed to high temperatures to damage and kill cancer cells, or to make cancer cells more sensitive to the effects of radiation and certain anticancer drugs. In the Academic Medical Centre in Amsterdam (the Netherlands), an external heating system has been developed using the AMC 70 MHz four antenna array [
8]. To evaluate the additional value of hyperthermia, a phase II study was initiated, combining radiochemotherapy plus deep regional hyperthermia (thermochemoradiation therapy = ThCR) as neoadjuvant treatment in operable oesophageal cancer patients.
Postoperative alterations in host immune functions after major surgical interventions have been extensively described and investigated [
5,
6,
13]. A previous study by our group, investigating the alterations in immune responses after limited transhiatal oesophagectomy (THO) and extended transthoracic oesophagectomy (TTO), demonstrated that both THO and TTO severely suppressed T helper type-1 (Th1) and type-2 (Th2) host immune responses. The extent of the surgical procedure had a differential immunosuppressive impact on Th2 cell activity (Th2 production of IL-4 and IL-13) but not on Th1 cell activity (Th1 production of IL-2 and IFN-γ). The two operations similarly impaired T helper type 1 cytokine production, which indicates that the two T helper pathways were down regulated through distinct mechanisms [
42]. The addition of preoperative thermochemoradiation therapy may suppress the immunological state of the patients even to a greater extent than oesophagectomy alone, but only limited data are available on immune responses after neoadjuvant therapy followed by surgery [
12,
20]. Furthermore, in the interval between neoadjuvant therapy and surgery recovery may take place, which is an important prerequisite for immunocompetence during and after surgery. The host recovery capacity and extent of recovery in the neoadjuvant setting is not well known.
Therefore, the objective of this study was to compare immunological parameters in patients with oesophageal cancer subjected to curative surgery with and without ThCR. Our analysis focused on the kinetics and recovery of lymphocyte subsets as well as stimulated Th-cell cytokine production that reflects the critical balance between Th1- and Th2-mediated immune responses.
Methods
Patients
Between January 2004 and January 2005, 32 patients with histologically proven cancer of the oesophagus or oesophagogastric junction consented to participate in the present study. All patients were scheduled for a potentially curative oesophagectomy by either a limited transhiatal or an extended transthoracic approach. Twenty patients underwent neoadjuvant therapy prior to surgery (neo patients) and 12 patients (control patients) underwent a potentially curative oesophagectomy without preoperative therapy. Patients younger than 75 years could participate in the prospective nonrandomised trial to undergo neoadjuvant therapy if there was less than 2 cm gastric involvement. Other inclusion criteria were written informed consent and mentally, physically, and geographically able to undergo treatment and follow-up. Exclusion criteria were diabetes; uncontrolled heart failure, hypertension, severe arrhythmia, and pacemaker; pre-existing myelopathy and/or polyneuropathy; previous radiotherapy and/or chemotherapy; age >75 years and <18 years.
Neoadjuvant therapy
Neoadjuvant therapy consisted of a 5-week schedule of ThCR. The schedule was as follows: chemotherapy, consisting of paclitaxel 50 mg/m
2 and carboplatin AUC = 2, by intravenous infusion on days 1, 8, 15, 22 and 29; radiotherapy consisting of a total of 41.4 Gy, given in 23 fractions of 1.8 Gy, 5 fractions per week, starting day 1 of the first cycle of chemotherapy; and hyperthermia with concurrent chemotherapy administration within 1 h after the radiotherapy treatment. The hyperthermia method has previously been described in detail by Albregts et al. [
1]. Four to 6 weeks after the end of neoadjuvant therapy, patients were scheduled for surgery.
Histopathological evaluation
All specimens were evaluated by an experienced gastrointestinal pathologist, in accordance with the criteria of the International Union Against Cancer, including stage, R classification and grade. Tumours were classified as histopathologically responding when less than 10% viable tumour tissue was found in the tumour bed; otherwise the tumour was classified as histopathologically non-responding [
44]. The group of non-responders was further divided into patients with partial response (10–50% viable residual tumour tissue), minimal response (>50% viable residual tumour tissue), and no change (absence of any regressive changes). In the group of responders, response was classified as complete (histological fibrosis with no viable residual tumour cells) and subtotal response (<10% viable residual tumour cells) [
21].
Blood sampling
Peripheral venous blood samples were obtained before neoadjuvant therapy, after 2 weeks of the start of neoadjuvant therapy, 1 day before surgery, on postoperative days 1, 3, 7, and 6 weeks after surgery (in total 7 sampling time points). In control patients, blood was drawn at the same time points except for the samples taken before and during neoadjuvant therapy (a total of 5 sampling time points).
Leucocytes and differential counts
The Beckman Coulter ACT Diff Analyser (Global Medical Instrumentation, Inc., Ramsey, MN, USA) was used for leucocyte and differential counts in peripheral venous blood samples, collected in ethylenediamine tetra-acetic acid (EDTA) tubes.
Flow cytometry
Peripheral venous blood samples were collected in EDTA collection tubes. The monoclonal antibodies used for immunophenotyping were from Becton Dickinson (San Jose, CA, USA). Multitest reagents were used in combination with True Count Tubes and FACS Lysing Solution (all Becton Dickinson), according to the instructions of the manufacturer. The antibody combination CD3CD8CD45CD4 or CD3CD16CD56CD45CD19 was applied: antibody CD45 for lymphocytes, CD3 for T lymphocytes, CD3CD4 for T helper lymphocytes, CD3CD8 for T cytotoxic lymphocytes, CD3CD16CD56 for natural killer cells and CD19 for B lymphocytes.
Flow cytometric analysis was performed on an FACS Calibur flow cytometer (Becton Dickinson) using the Multiset software package (Becton Dickinson). A minimum of 2,000 and 5,000 lymphocytes were measured for T cell and for NK and B cell analysis, respectively.
Whole blood cultures
Whole heparin blood was diluted 1:10 in Iscove’s modified Dulbecco’s medium (IMDM; Boehringer Ingelheim, Alkmaar, The Netherlands), supplemented with 0.1% faetal calf serum, penicillin (100 IU/ml), streptomycin (100 μg/ml), and 15 IU/ml sodium heparin (Leo Pharmaceutical Products, Weesp, The Netherlands) and was cultured in triplicate in flat bottom 200-μl wells (Nunc, Roskilde, Denmark). In order to assay the stimulation of cytokines, diluted whole blood was stimulated with anti-CD3 (CLB-T3/4.E, 100 ng/ml) in concert with anti-CD28 (CLB-CD28/1, 1 μg/ml; Sanquin, Amsterdam, The Netherlands) for cross-linking of T cell CD3 and CD28 receptors. Supernatants were harvested after 24 h (for IL-2) and after 72 h (for IFN-γ, IL-4, and IL-13) and then stored at −80°C until analyses.
Cytokine assays
Supernatant cytokine levels were measured using appropriate combinations of singleplex assays from Bio-Rad (Veenendaal, The Netherlands). Supernatant was diluted 1 in 4, using Bio-Plex human serum diluent, and fluorescence signals were analysed on the Bio-Plex reader after calibration with the high calibration setting. Cytokine production per T cell was calculated by dividing the amount of each cytokine (pg/ml) by the number of CD3+ cells.
Statistical analysis
All data were analysed using Statistical Software Package version 12.0.2 (SPSS Inc., Chicago, IL, USA) for Windows XP. Comparisons between groups were made using the Mann–Whitney test (for continuous variables). To determine the effect of neoadjuvant therapy on immune responses, the Wilcoxon rank test was applied to compare values 2 weeks after neoadjuvant therapy to those at baseline. To determine the effect of surgery on immune responses, the Wilcoxon rank test was applied to compare values of days after operation to those at baseline within the group. Between-group differences over time were analysed by a general linear model for repeated measures. By checking, residuals were approximately normally distributed. Standard descriptive analysis was performed for values at baseline. In figures, data are presented as mean values ± one standard error of the mean. Statistical significance was defined as P < 0.05 (two-sided).
Discussion
Major surgery is known to be associated with immune suppression. Several studies, investigating alterations in immune defence after oesophageal resection, have shown that oesophagectomy depresses host immune response [
41,
42]. However, little is known about the immunological effects of preoperative neoadjuvant treatment in these patients [
40].
The present study describes a broad spectrum of both numerical and functional immunological parameters in a series of surgical patients, some of whom received preoperative treatment, while others were directly operated on. The study population was uniform in that all patients had potentially curable disease based on preoperative investigations, and clinical parameters were comparable in both groups.
In brief, patients who received neoadjuvant thermochemoradiation therapy demonstrated a significant decrease in the number of granulocytes and lymphocyte subsets. T cell cytokine release was decreased after 2 weeks of neoadjuvant therapy, while production per T helper cell remained the same or showed a modest increase. CD8+ (cytotoxic) T cells but not CD4+ (helper) T cells recovered in the period between neoadjuvant therapy and operation. CD4+ T (helper) cell cytokine production levels were almost recovered, preoperatively.
All lymphocyte subsets that had not recovered after neoadjuvant therapy (i.e., CD4+ T cells, NK and B cells but not CD8+ T cells) had persistent lower counts as compared to control patients (operated without neoadjuvant therapy) at 1 week postoperatively.
Hyperthermia causes a transient stimulation of the cell-mediated immune response [
38]. Recent research has shown that, when cancer cells are heated to approximately 41.6°C, they form highly specific protein structures on their surface known as heat-shock proteins [
27]. These proteins activate the patient’s own NK cells to attack the cancer cells and stimulate the immune system [
27]. Studies, evaluating the immune response after chemotherapy and/or radiation therapy, showed profound negative effects on the immune system [
18,
22,
29,
36,
43]. In the present study, the neo patients underwent a combination of chemotherapy, radiotherapy and hyperthermia. The cumulative effects of this combination therapy led to a reduction of the white blood cell count which resulted in an overall reduced capacity to produce cytokines after 2 weeks of therapy. Apparently, the immunosuppressive impact of chemoradiation surpasses the immunostimulatory effect of hyperthermia.
In patients with sepsis, the loss of lymphocytes, contributing to immunosuppression by decreasing the number of available immune cells, is a well-known phenomenon. However, relatively recent work implicates that, not only a decrease in the number of effector cells, but also lymphocyte apoptosis plays a potential factor in the immunosuppression in sepsis [
16]. The apoptotic cells appear to have important adverse effects on immune function by actively suppressing the inflammatory response [
15]. It is unknown at present whether lymphocyte apoptosis plays a role in thermochemoradiation-induced immunosuppression, and may be an important subject for further study.
The differential recovery of CD8+ cells in contrast to CD4+ cells leading to a prolonged T cell imbalance after thermochemoradiation therapy has been described by others after intensive chemotherapy and after peripheral blood stem cell transplantation [
19,
31]. As most peripheral CD4+ en CD8+ T cells share a common primary developmental pathway until relatively late in their development, this significant difference in the rate of recovery is surprising. In short, T cell precursors move from the bone marrow to the thymus, and develop to T cells. In the thymus various T cell precursors can be distinguished by the expression of specific cell surface markers: first TCR (T cell receptor), second CD3 (which serves as the signal transduction component of TCR), and finally CD4/CD8. It could be hypothesised that the regeneration of both CD4+ and CD8+ T cells occurs exclusively via thymic dependent pathways [
19]. However, there is no existing evidence to suggest that a suboptimal thymic regenerative capacity would preferentially affect CD4+ populations, which could explain the differential recovery rate [
3,
32,
37]. Alternatively, extrathymic generation of CD8+ cells from haematopoietic precursors could be involved and/or peripheral expansion of mature CD8+ T cells, which remain in the secondary lymphoid tissues after completion of chemotherapy. Mackall et al. [
19] suggested that extrathymic lymphopoiesis is involved, as in their study the predominance of CD8+ cells lacked CD28 expression. This lack of CD28 expression has been associated with impaired thymic function in HIV infection [
4], post-bone marrow transplantation [
2,
24], and aged hosts [
34]. Expansions of CD8+CD57+ cells have also been described in combination with restricted T cell repertoire diversity [
11]. Thus, the relatively rapid regeneration of CD8+ cells compared to the slow CD4+ regeneration might be due to thymic independent pathways that efficiently and selectively induce regeneration of distinct CD8+ subsets.
Furthermore, differences in sensitivity of the factors that regulate homoeostatic proliferation of memory CD8+ T cells as compared to memory CD4+ T cells may explain the slower recovery of the CD4+ T cell-population from thermochemoradiation. Homoeostatic proliferation of CD8+ T cells seems to depend on IL-15 and IL-7, whereas these cytokines are probably not required for homoeostatic proliferation of CD4+ T cells [
39].
Further work is necessary to be able to specify the origin of regenerated CD8+ cells and either reject or sustain the explanation of biological distinctions between CD8+ and CD4+ T cell regenerative pathways. It may be of clinical importance to determine if a predominance of CD8+ cells exists of CD8+CD28
- and CD8+CD57+ subsets, since a limited effectiveness of these populations has been described in vivo [
30]. A significant depression in cell-mediated immunity in conjunction with possible alterations of CD8+ T cell subsets in neo patients as compared to control patients may have implications for the host defence against foreign pathogens as well as in the immunological control of disseminated tumour cells.
Although the role of the immune response in controlling tumour growth and cancer recurrence is controversial, there is evidence in oesophageal cancer suggesting that T cell infiltrates have a beneficial prognostic impact [
14]. It was shown that CD8+ T cell infiltrations within the tumour specimen have a favourable outcome [
35]. And a recent research in patients with colorectal cancers indicated that type, density and location of immune cells within tumour tissues were a better predictor of patient survival than the current methods to stage colorectal cancers [
10,
28]. Interestingly, colorectal tumours from patients without recurrence had higher immune cell densities of CD3+ cells within the tumour and at the infiltrating tumour margin. Among CD4+ T lymphocytes, Th1 and Th2 phenotypes can be distinguished [
25]. Activation of these two functionally distinct subsets of mature T helper cells is of great importance for an effective immune response against infection. Th1 cells are mainly characterised by secretion of IL-2 and IFN-γ and induce cell-mediated immune responses. Th2 cells preferentially secrete the cytokines IL-4 and IL-13 which favour humoural immune responses [
7,
33]. Previously published results from our group showed that oesophagectomy severely depressed host immune responses. In the present study, neoadjuvant therapy caused both reduced T cell counts and decreased absolute levels of both Th1 and Th2 derived cytokines. This is in line with previous investigations, studying alterations in immune defence after neoadjuvant treatment [
12]. However, in the present study it was shown that after neoadjuvant therapy production of the Th1 and Th2 cytokines per T helper cell was not decreased and for IL-13 even moderately increased, indicating that reduced cell counts rather than functional deficits contributed to the post-therapy reduction of T cell cytokine production.
At present, novel therapeutic strategies in oesophageal cancer mainly focus on applying new chemotherapeutic drugs, varying drug dosages, optimising the dose of radiotherapy or on adding hyperthermia to optimise long-term survival. However, the role of the host immunocompetence in the inhibition of disease progression and the way it is altered by the various new therapeutic strategies have only been partially investigated. The present study shows significant disturbances of host cellular immunity induced by preoperative thermochemoradiation compared to surgery alone in oesophageal cancer patients. This neoadjuvant therapy caused reduced T, NK and B cell counts, reduced overall Th1 and Th2 cytokine production, and differential recovery of T cytotoxic and T helper cells, leading to prolonged T cell imbalance that extends beyond the time of surgery. However, the study group was to small to conclude on the basis of the immune response whether to stop or continue neoadjuvant therapy, although it indicates that for individual patients it is worthwhile monitoring their immune status, having a vigilant attitude toward emerging infections. The oncological consequences of these immunological changes need further investigation, as recovery of T helper cytokine production towards surgery was less impaired than T helper cell counts.