The degree of immunosuppression in patients with haematological malignancies treated with chemotherapy is routinely measured as number of circulating cells (preferable neutrophils) in peripheral blood. A parallel decline in the number of T cells is expected, but a possible alteration in their functionality has been less well explored. The ability of T cells to secrete more than one cytokine simultaneously is known to indicate protective immunity. The aim of this study was to determine whether the function of circulating T cells is altered in patients with chemotherapy-induced neutropenia.
Design, setting and subjects
In this cross-sectional study, we used the FluoroSpot assay to investigate the proportion of T cells secreting either interferon-γ or interleukin-2, or both cytokines simultaneously, after anti-CD3 stimulation. Peripheral blood mononuclear cells from 53 adult patients with chemotherapy-induced neutropenia and 20 healthy individuals were investigated.
There were significantly fewer T cells secreting interferon-γ in patients with neutropenia compared with healthy control subjects (P = 0.02), but the difference was greatest for dual cytokine-secreting T cells (P = 0.001). Furthermore, the amount of secreted cytokine per T cell appeared to be reduced in patients, compared with control subjects.
Our results suggest that the functionality of T cells is altered in patients with haematological malignancies with chemotherapy-induced neutropenia. In parallel with a decline in T cell count, this may further increase the risk of severe infections.
In patients treated for haematological malignancies, the relationship between the risk of infection and the degree of immunosuppression is well known , and neutropenia is commonly used as a measure of immunosuppression . However, neutropenia may be accompanied by a parallel decline in other immunological cell types such as T cells that play an important role in the adaptive immune response to infectious diseases . The potential effects on T cell function of immunosuppressive treatment for haematological malignancies have been evaluated with contradictory results [4-6]. In a recent study by Fuhrmann et al.,  immunosuppressive drugs including cyclosporine A and tacrolimus reduced cytomegalovirus (CMV)-specific T cell ‘polyfunctionality’ defined as simultaneous expression of interferon-γ (IFNγ), interleukin-2 (IL-2) and tumour necrosis factor-α. However, the single IFNγ responses were not affected. The ability of T cells to respond to specific antigens with a combination of cytokines rather than only one specific type has been associated with enhanced protective immunity against both viruses and bacteria [8-10]. For example, various recent studies have demonstrated a correlation between the number of HIV-specific CD4+ T cells that concomitantly secrete both IFNγ and IL-2 and a good clinical outcome in HIV infection, whereas no difference in the number of specific T cells secreting either IFNγ or IL-2 alone was observed between groups with a different clinical prognosis [11, 12]. The type of functional responses is a measure of quality, and the number of responsive cells is a measure of magnitude; both of these are important in evaluating the efficacy of immune responses . Increased knowledge and ability to monitor the preserved T cell function in immunocompromised patients are important for design of viral surveillance programmes as well as potential future immunotherapy, such as leukaemia-specific cytotoxic T cells.
In this study, we compared the cytokine response pattern from peripheral blood mononuclear cells (PBMCs) in healthy individuals and from patients with chemotherapy-induced neutropenia. A FluoroSpot assay was used for fluorescent detection and quantification of T cells secreting IFNγ alone, IL-2 alone and IFNγ/IL-2 simultaneously .
Materials and methods
Study population and ethical statement
Blood samples were collected from patients with haematological malignancies and chemotherapy-induced neutropenia (absolute neutrophil count ≤0.5 × 109 L−1) included in a cross-sectional study of viral infections . Thirty-five of the samples were from randomly selected patients with febrile neutropenia (fever defined as a single temperature reading ≥38.5 °C or temperature >38.0 °C for 1 h), and 25 were from randomly selected patients with neutropenia but without fever or signs of infection. Blood samples were also collected from 20 healthy volunteers specifically for this study.
All study subjects provided written informed consent before enrolment. The study was carried out at the Karolinska University Hospital and was approved by The Regional Ethical Review Board in Stockholm.
Blood samples were collected in two lithium heparin tubes (8 mL each) for leucocyte separation using density gradient separation (Ficoll-Paque™ PLUS; GE healthcare Life Sciences, Uppsala, Sweden). Additional peripheral blood was collected from patients with neutropenia for extensive viral diagnostic assessment by real-time PCR and T cell counting by flow cytometry, as previously described in detail . Clinical data were extracted from the patients' medical records.
After density centrifugation, PBMCs were collected, washed twice in phosphate-buffered saline (PBS) and resuspended in freezing medium containing 90% heat-inactivated foetal calf serum and 10% DMSO. Aliquots of this cell suspension were then transferred into cryovials, placed into a freezing container (‘Mr. Frosty’, Thermo Fisher Scientific, Waltham, MA, USA) and stored at −80 °C. After 20 h, the samples were stored at −155 °C.
On the day of the FluoroSpot assay, vials of cryopreserved PBMCs were thawed in a water bath at 37 °C until a small ice-cube remained. The cell suspension was transferred into a polypropylene tube and slowly mixed with 7 mL cell culture medium (RPMI 1640 containing 10% foetal calf serum, 1 mmol L−1 glutamine, 100 units mL−1 penicillin, 100 μg mL−1 streptomycin and 0.5 mmol L−1 HEPES; all from Invitrogen life technologies, Carlsbad, CA, USA). After washing the cells twice at 200 g for 10 min in cell culture medium, the PBMCs were counted and analysed for viability using the Guava Viacount assay and the EasyCyte Mini System (Millipore, Solna, Sweden). Based on cell counts, after excluding both dead and apoptotic cells, the PBMCs were diluted in cell culture medium to a concentration of 1 × 106 cells mL−1.
T cell responses to anti-CD3 were evaluated at the single-cell level using an IFNγ/IL-2 FluoroSpot assay in accordance with the manufacturer's guidelines (Mabtech AB, Nacka Strand, Sweden) . Thawed PBMCs (50 000 cells/well) were added to precoated (IFNγ/IL-2) FluoroSpot plates and incubated for 20 h in the presence of anti-CD3 monoclonal antibody (mAb) (CD3-2, Mabtech AB, diluted 1 : 1000) and anti-CD28 mAb (CD28-A, Mabtech AB, 0.1 μg mL−1). In control wells, PBMCs were incubated with the anti-CD28 mAb alone. The following day, plates were washed in PBS and captured cytokines were detected using a combination of antibodies conjugated to FITC (IFNγ) and biotin (IL-2) with the subsequent addition of anti-FITC antibody conjugated to a green fluorochrome and streptavidin conjugated to a red fluorochrome. The numbers of spot-forming units were determined using an iSpot reader system (AID, Strassberg, Germany) with separate filters for FITC and Cy3. Cells secreting both IFNγ and IL-2 were determined based on spots with the same central point in an image overlay of FITC and Cy3 images.
For univariate analyses comparing two groups, Fisher's exact test and the Mann–Whitney U-test were used for categorical and continuous data, respectively. All tests were two-sided, and P < 0.05 was considered statistically significant. Prism 5.00 (GraphPad Software, San Diego, CA, USA) for Windows software was used for all analyses.
The group of healthy volunteers included 10 women and 10 men, whereas the group of patients with neutropenia included 21 women and 32 men (P = 0.44). With a median age of 46 years (range 27–69), the healthy individuals were younger than the patients with neutropenia (median 60 years, range 25–86, P < 0.05). The characteristics of the neutropenic patients are shown in Table 1.
Table 1. Characteristics of patients with neutropenia (n = 53)
Amongst patients with febrile neutropenia, 10 also showed concomitant viraemia: polyoma BK virus (n = 5), CMV (n = 4) or polymicrobial viraemia with CMV and parvovirus B19 (n = 1). Viral copy numbers were <1000 copies mL−1 except in two cases of polyoma BK virus (37 850 and 7050 copies mL−1) and the only case of parvovirus B19 (1220 copies mL−1).
T cell counts
The CD4+ and CD8+ T cell counts of the healthy volunteers were within the normal reference range of the Karolinska University Laboratory at Karolinska University Hospital, whereas these cell counts were significantly suppressed in patients with neutropenia (Fig. 1 and Table 2). The patients with febrile neutropenia had lower cell counts than those included in the study in an afebrile state, but the differences were not statistically significant (Fig. 1 and Table 2). Furthermore, no statistically significant differences were detected between the febrile patients with (n = 10) and without (n = 18) viraemia in terms of both CD4+ and CD8+ T cell counts (P = 0.22 and P = 0.76, respectively; Fig. 1).
Table 2. T cell counts and number of cytokine-secreting cells
Healthy controls (n = 20)
Patients with neutropenia (n = 53)
Afebrile (n = 25)
Febrile (n = 28)
Cell counts (×109 L−1 peripheral blood) and numbers of secreting cells (per 50 000 T cells) are presented as median (range).
T cell counts
Number of T cells secreting IFNγ and IL-2
There were fewer T cells secreting cytokines after anti-CD3 stimulation in patients with neutropenia compared with the healthy individuals (Fig. 2 and Table 2). However, the difference reached statistical significance only for IFNγ- (P = 0.02) and dual cytokine-secreting T cells (P = 0.001). The amount of cytokines per secreting cell was found to be higher in the healthy individuals compared with the patients with neutropenia (data not shown). The difference in fluorescence intensity between these groups is illustrated in Fig. 3. The number of cytokine-secreting T cells did not differ between afebrile and febrile patients (Fig. 2 and Table 2). Although not statistically significant, the number of cytokine-secreting cells tended to be higher in febrile patients with viraemia compared with nonviraemic patients (Fig. 2). The difference was greatest for T cells secreting IL-2 alone (median 681 vs. 65, P = 0.11), followed by dual cytokine-secreting (median 91 vs. 24, P = 0.22) and IFNγ-secreting T cells (median 1007 vs. 579, P = 0.20).
In this study, we compared cytokine responses in PBMCs from healthy individuals and patients with haematological malignancies with chemotherapy-induced neutropenia. A FluoroSpot assay was used to estimate the number of T cells secreting either IFNγ or IL-2, or both simultaneously, after anti-CD3 stimulation. The responses were generally weaker in the patients with neutropenia, and the difference was most prominent for the number of dual IFNγ-/IL-2-secreting cells.
Previous studies have indicated that the function of T cells is maintained during chemotherapy-induced neutropenia in children treated for acute leukaemia [4, 5]. No major abnormalities were observed in these patients, and their memory T cell compartments were relatively well preserved. In allogeneic transplant patients, however, activation of cytotoxic T cells was impaired . Fuhrmann et al. recently concluded that cyclosporin A and tacrolimus reduced T cell polyfunctionality but not single IFNγ responses against CMV . In the present study, not only was the T cell count reduced, but also the proportion of T cells responding to anti-CD3 stimulation and in particular the number of dual cytokine-secreting T cells decreased too. Furthermore, the amount of cytokine secreted per T cell seemed to be higher in the healthy individuals compared with the immunocompromised patients. Thus, our data suggest that there is an overall decline in circulating T cells in patients with chemotherapy-induced neutropenia. In addition, the proportion of these cells responding to anti-CD3 was reduced and the amount of cytokine secreted by responding cells appeared to be less. Finally, dual-secreting T cells, which have been associated with protective immunity [11, 12], are most affected suggesting that both the magnitude and quality of the immunological response are compromised.
When chronic infection with CMV is used as a model of protracted antigen exposure and antigen persistence with low antigen load, the CD4+ T cell response consists of an equally distributed mixture of cells with an IFNγ, IL-2 and dual IFNγ/IL-2 signature [17, 18]. Under these circumstances, the CD8+ T cells consist of approximately 80% and 20% IFNγ- and dual cytokine-producing cells, respectively. Upon primary infection or a noncontrolled chronic infection, however, the pattern is different with a dominant IFNγ signature in both CD4+ and CD8+ T cells. We measured the response from the total population of T cells and found a mixture of these three subsets. In both healthy controls and patients with neutropenia, there was a predominance of single IFNγ responses representing approximately two-thirds of the responding cells and less than 10% were dual cytokine-secreting T cells. Patients with viraemia had similar numbers of responding T cells compared with the other patients. Although the T cells were polyclonally stimulated, it is possible that reactivation would be more likely in patients with low numbers of responding cells. However, the viral titres detected were low, and there were no reports of development of viral disease, which reflects the fact that viral reactivation was under control in these patients.
To minimize undesirable cytotoxicity, doses of chemotherapy may in future be reduced by accompanying immunotherapy, such as leukaemia-specific cytotoxic T cells. Although the T cell responses in our chemotherapy-induced patients with neutropenia were altered both in magnitude and quality, patients with viral reactivation had similar T cell capacity to other patients and seemed to be able to control the infection. The presence of a T cell compartment rich in memory T cells, which has been observed previously in neutropenic patients, could explain this control . It is clear that a more specific immunological selection and evaluation for each potential immunotherapy must be considered in future studies, but our results suggest that the T cell compartments in patients with neutropenia might be sufficiently intact for immunotherapeutic strategies.
Our method could be suitable to assess the functional capacity of circulating virus-specific T cells in allogeneic haematopoietic stem cell transplant recipients. Recently, dual IFNγ-/IL-2-producing CD8+ T cells but not single IFNγ-producing CD8+ T cells were associated with control of CMV reactivation after transplantation , and Gayoso et al.  have suggested that a cheap, reliable and simple method for this immune monitoring is needed. The low number of circulating cells after treatment-induced cytopenia poses a technical problem for investigation of the function of the T cell system. In many previous studies, this problem has been solved by investigating monoclonal in vitro expanded T cell populations, but only clonogenic subsets can be examined with such an approach. We were able to use PBMCs from a total of 16 mL whole blood and could examine the response from the whole T cell repertoire. Furthermore, the use of the FluoroSpot technique allowed us to measure the accumulated T cell secretion of either one or two cytokines.
A number of limitations of this study should be considered. First, we did not measure proliferative capacity or cytotoxicity. However, it has been suggested that the dominance of either IFNγ, IL-2 or multifunctional (IFNγ and IL-2) responses may define the main functional patterns of T cells . Secondly, we did not separate CD4+ and CD8+ T cells, and thus, the results could have been influenced by a predominance of one cell type. There were, however, similar proportions of these T cells in the different groups, and both CD4+ and CD8+ T cells have a mixed response to low levels of virus in peripheral blood and a dominating IFNγ response upon uncontrolled infection . Finally, Han et al.  demonstrated that T cells initiate the release of cytokines at different time-points after stimulation; most of the cells are initially monofunctional, and the multifunctional state is short-lived. Thus, dynamic monitoring was suggested to better distinguish subpopulations.
In conclusion, we were able to demonstrate a decreased anti-CD3-induced Th1 pro-inflammatory response in patients with chemotherapy-induced neutropenia. In particular, in addition to the profound decline in T cell counts, we found a reduced number of IFNγ-/IL-2-secreting T cells, which have been associated with protective immunity. This potentially further increases the risk of severe infections.
We are grateful to all study participants and thank the nurses and physicians at the Haematology Centre, and staff at the Accident and Emergency Department and the Department of Clinical Microbiology at the Karolinska University Hospital, for their assistance. This study was supported by grants from the Stockholm County Council /Karolinska Institutet, the Swedish Society of Medicine, the Clas Groschinsky Foundation, the Swedish Childhood Cancer Foundation and the Karolinska Institutet Infection Network. The funding bodies had no role in the study design, data collection and analysis, decision to publish or preparation of the manuscript.
Conflict of interest statement
CS is employed by Mabtech, the manufacturer of the FluoroSpot assay. All other authors declare that they have no conflict of interest.
LÖ, CS, KB, TT and ON designed the experiments; LÖ, CS and ON performed the experiments; LÖ, CS, MW and ON analysed the data; CS, KB and TT contributed reagents/materials/analysis tools; and all authors contributed to writing the manuscript.