Major Surgery Suppresses Maximal Production of Helper T-Cell Type 1 Cytokines Without Potentiating the Release of Helper T-Cell Type 2 Cytokines FREE

DURING THE past 2 decades, clinicians and researchers have recognized that surgery results in an impaired immune response in the early postoperative period¹ and that this phenomenon may contribute to infectious complications² and cancer metastasis.³ The immunosuppressive effect of surgery has been variously attributed to serum factors,⁴ stress hormones,^4- 5 the activation of suppressor cells,^5- 6 decreased secretion of the cytokines interferon γ (IFN-γ) and interleukin 2 (IL-2),^7- 8 and, most recently, to a shift in the balance of helper T-cell type 1 (T_H1) and type 2 (T_H2) cytokines from peripheral blood mononuclear cells.^9- 10 A better understanding of perioperative immune function might enable clinicians to stratify patients as immunologically high risk before surgery and take appropriate preventive action or delay surgery. Alternatively, patients demonstrating substantial postoperative immune suppression could be considered for additional prophylactic treatment aimed at reducing the risk of infectious complication.

The present study used recently developed flow cytometry techniques for direct intracellular cytokine measurements along with current concepts of lymphocyte immunomodulation to arrive at a more precise understanding of the effects of surgical injury on immune function. The hypothesis was that surgical trauma suppresses cytokines through an increase in the production of T_H2 cytokines.

This study consisted of a cohort of 18 patients in the Veterans Affairs and US Department of Defense who underwent elective aortic surgery (n=10) or carotid endarterectomy (n=8). The study was approved by the Institutional Review Board at David Grant Medical Center, Travis Air Force Base, Calif, and all patients signed informed consent forms prior to participation in the study. Entry criteria excluded patients with diseases known to affect the immune system, such as acquired immunodeficiency syndrome, chronic hepatitis B and C, known active malignancy, and the use of immunosuppressive medications. Samples of peripheral venous blood (20 mL in EDTA for flow cytometry and 7-mL clot tube for serum assays) were obtained by venipuncture from each patient no more than 6 hours preoperatively and every morning after surgery until discharge from the hospital or for a maximum of 5 days. A group of 17 healthy volunteers were studied as reference subjects.

T-cell subsets and their intracellular cytokine production were identified by multiparametric flow cytometry.^11- 13 Peripheral blood mononuclear cells were isolated from 20 mL of anticoagulated (EDTA) venous blood by density centrifugation and incubated at 37°C in a 5% carbon dioxide incubator for 4 hours in RPMI (Roswell Park Memorial Institute; Sigma-Aldrich Corp, St Louis, Mo) medium containing 10-µg/mL brefeldin A (Epicentre Technologies, Madison, Wis) to dissaggregate the Golgi complex and enable the intracellular accumulation of proteins.¹⁴ During the incubation period, half the cells were separately stimulated with 20-ng/mL phorbolmystrateacetate (Sigma-Aldrich Corp) and 1-µg/mL ionomycin (Sigma-Aldrich Corp) to assess submaximal cytokine production, while the other half remained unstimulated. After incubation, cell suspensions were divided into 18 tubes with approximately 1×10⁶ cells per tube corresponding to cell type (CD4⁺, CD8⁺, or γδ⁺), cytokine combination (isotype control, IFN-γ/IL-4, or IL-2/IL-10), and stimulation category (unstimulated or stimulated). Cells were resuspended in buffer (phosphate-buffered saline plus 0.1% sodium azide plus 1% bovine serum albumin plus 5-mmol EDTA) and incubated for 15 minutes at 25°C with a 1:5 dilution of Tricolor-conjugated (a tandem conjugate of phycoerythrin and Cy-5; Caltag Laboratories, San Francisco, Calif) anti-γδ T-cell receptor, anti-CD4, or anti-CD8 monoclonal antibodies for phenotypic identification. After washing in phosphate-buffered saline, cells were fixed in 100 µL of solution A (cell permeabilization fixation; Caltag Laboratories) for 15 minutes at 25°C and washed again in phosphate-buffered saline. Intracellular cytokines were identified by resuspension in 50 µL of the permeabilization solution B (Caltag Laboratories) containing a 1:500 dilution of dinitrophenyl-conjugated anti–IFN-γ and a 1:100 dilution of biotin-conjugated anti–IL-4; or dinitrophenyl-conjugated anti–IL-10 and biotin-conjugated anti–IL-2; or dinitrophenyl-conjugated anti-mouse IgG1 and biotin-conjugated anti-mouse IgG2 for 15 minutes at 25°C. Specific intracellular cytokine identification was achieved by subsequent incubation (Strepavidin Red 670; Pharmingen Inc, San Diego, Calif) and fluorescein isothiocyanate–conjugated anti-dinitrophenyl antibody (Molecular Probes Inc, Eugene, Ore) for 15 minutes at 25°C. Finally, cells were resuspended in 0.5 mL of cytoflow buffer for analysis. Surface and intracellular fluorescence was measured using a flow cytometer (FACScan; Becton Dickinson Immunocytometry Systems, San Jose, Calif) and fluorescence data were analyzed with software (CellQuest; Becton Dickinson Immunocytometry Systems). After standard calibration of the flow cytometer, data were recorded until 5000 cells were acquired or the cell suspension volume of 0.5 mL was exhausted. The minimum number of cells was lowered to 2000 cells for γδ⁺ T cells since these cells are much scarcer in the peripheral blood. Intracellular cytokine detection was performed by double gating each cell surface marker and a lymphocyte gate and measuring fluorescence channel 1 (fluorescein isothiocyanate) and fluorescence channel 3 (Strepavidin Red 670) fluorescence intensity. The data recorded included the total number of acquired cells, the percentage of total cells gated for each subset, and the number of cells positive for fluorescence channel 1 and fluorescence channel 3 internal fluorescence. Background fluorescence was measured using isotype-matched fluorochrome-conjugated IgG and arithmetically subtracted from the measured intracellular cytokine fluorescence to obtain the net detectable intracellular fluorescence. The average background fluorescence measured was less than 1%. Reference assays using fixed, but not permeabilized, cells demonstrated a reduction in the intracellular detection of IL-2 and IFN-γ by 99% and 98%, respectively. Preincubation with recombinant human cytokines before intracellular staining reduced intracellular detection of IL-2 and IFN-γ by 99% and 98%, respectively. Serum IL-6 levels were measured by enzyme-linked immunosorbent assay by the Clinical Investigation Facility at David Grant Medical Center.

A complete data set in both surgical groups was obtained for the preoperative point and the first 2 postoperative days. These data were analyzed using 2-way repeated-measures analysis of variance and Tukey post hoc test. Patient characteristics were compared using χ² analysis for proportional data and the Kolmogorov-Smirnov test for unpaired comparisons that were not normally distributed. All statistical analyses were performed with statistical software (STATISTICA; StatSoft, Tulsa, Okla) for Windows (version 5.0; Microsoft Corporation, Redmond, Wash).

The patient characteristics for the 2 surgical groups are listed in Table 1. Complete blood cell counts (Figure 1) demonstrated significant postoperative increases in the numbers of neutrophils in both groups compared with preoperative levels. Absolute lymphocyte counts decreased in both groups compared with preoperative levels. The postoperative platelet count decreased significantly in the aortic surgery group but not in the carotid surgery group. Postoperative serum IL-6 levels were significantly elevated in the aortic surgery group only (Figure 1).

Unstimulated intracellular cytokine production was not detectable above nonspecific background activity in any T-cell subset. The stimulated production of the intracellular T_H1 cytokines IFN-γ and IL-2 was significantly decreased in the postoperative period in all T-cell subsets and in both surgical groups compared with preoperative values (Figure 2, Figure 3, and Figure 4, C and D). There were no increases in T_H2 cytokine production after surgery. Small preoperative differences (<1.5%) in the stimulated production of IL-4 in CD4⁺ and CD8⁺ cells and IL-10 in γδ T cells between surgery groups were at the threshold of nonspecific background activity, although the differences were statistically significant (Figure 2, Figure 3, and Figure 4, A and B). There were no significant differences in preoperative or postoperative stimulated intracellular cytokine production between patients when grouped by operation type, preoperative cardiovascular risk factors (diabetes mellitus, history of coronary artery disease, and smoking), the presence of homologous or autologous transfusion, intraoperative blood loss, operating time, hospital stay, or the development of infectious complications.

There were no significant postoperative changes in unstimulated production of T_H1 or T_H2 cytokines after major vascular operations. These results suggest that major surgery does not trigger a systemic antigen-specific immune response. In contrast, stimulated (potential) production of T_H1 cytokines in CD4⁺, CD8⁺, and γδ⁺ T cells significantly decreased in the postoperative period (P<.05). This effect persisted through the second postoperative day and was independent of the level of the type of operation performed. The decrease in stimulated production of T_H1 cytokines was not accompanied by a simultaneous postoperative increase in production of T_H2 cytokines. There were no differences in the cytokine responses when patients were stratified by preoperative medical diagnosis or by the presence of postoperative complications, although the small sample size limits the power of this observation.

The results of the present study are consistent with previous reports^15- 20 that demonstrated that surgical trauma results in a depression of in vitro lymphocyte blastogenesis and proliferation during the early postoperative period. This effect has been attributed to stress-mediated decreases in the production of the T_H cytokines IL-2 and IFN-γ^8,21- 23 when measured in bulk supernatants from in vitro cultures of peripheral blood mononuclear cells. More recently, immunosuppression in human disease has been attributed to a shift in cytokine production from a T_H1(IL-2 and IFN-γ) to a T_H2 (IL-4 and IL-10) pattern.²⁴ The present study confirms a specific negative effect of surgical injury on the production of T_H1 cytokines without a simultaneous increase in the production of T_H2 cytokines. These findings differ from recent reports by Decker et al^9- 10 that suggested that postoperative suppression of T_H1 cytokine responses is associated with an increase in T_H2 cytokine production.

One explanation for the lack of an observed T_H2 cytokine response in the present study may be a high threshold of detection resulting from the use of a 2-step staining technique that might mask the expected 1% to 3% of T cells that produce IL-4 in response to phorbolmystrateacetate and ionomycin stimulation.¹¹ The determination of the exact patterns of postoperative cytokine production is also difficult because of the uncertainty of the precise timing of potential T_H2 responses after injury. The results of animal studies demonstrated that the T_H2 cytokine IL-4 is detectable in splenocytes within 24 hours of traumatic injury,²⁵ whereas the same cytokine is not measurable in splenic and peritoneal cavity T cells until 5 to 8 days after intraperitoneal infection with Nippostrongylus brasiliensis, a potent inducer of T_H2.¹⁴ Therefore, it is possible that the first cytokine measurement at 24 hours postoperatively was too late to detect an early T_H2 response, which was reported by others.¹⁰ It is also plausible that T_H2 responses developed in patients after the last measurement on postoperative day 2. However, a review of the limited data obtained in the present study between postoperative days 2 and 5 demonstrated no T_H2 cytokine production. Finally, cytokine production in the present study was only measured in T cells, not in other antigen-presenting cells or natural killer cells. These other cell types have been reported to produce T_H2 cytokines²⁴ in humans. Thus, T_H2 cytokine production by cells other than T cells may explain other reports¹⁰ of postoperative elevations of T_H2 cytokine levels not seen in the present study. One study²⁴ suggested, however, that it would be unlikely to observe T_H2 cytokine responses in humans that do not involve CD4⁺ T cells.

The observation that the degree of postoperative suppression of T_H1 cytokine responses was the same after carotid endarterectomy and aortic surgery was surprising. This finding suggests the occurrence of a generalized suppression of T-cell function during the postoperative period that may be independent of the degree of tissue injury. Such a generalized immune suppression following surgical injury may change the immunological significance of what has been defined as major and minor surgery. An unexpected finding was the large variance in preoperative stimulated intracellular cytokine levels in the aortic surgery group; this variance was not present in the carotid endarterectomy group. The difference in preoperative immune reactivity between groups may reflect differences in the patient population, underlying vascular disease, or other unmeasured confounding factors. The potential correlation of decreased preoperative immune function in patients undergoing aortic surgery needs to be further investigated with respect to the risk of postoperative infectious complications.

The initial cytokine milieu is one of the main factors determining T_H1 and T_H2 differentiation in T cells. γδ T cells are an epithelial-associated lymphocyte subset that respond early to infection and cell injury.^26- 27 These specialized lymphocytes can release T_H1 and T_H2 cytokines, which are thought to play a role in modulating the functional differentiation of the T-cell response.¹⁴ The present study demonstrates that γδ T cells are vigorous producers of IFN-γ, a finding reported by other investigators.^14,28- 29 The finding that the responses of γδ T-cell T_H1 cytokines are significantly down-regulated after surgery merits further investigation in patients who have undergone surgery because γδ T cells play an important role in the epithelial barrier to bacterial infections,³⁰ in neutrophil recruitment^26,31 and the process of epithelial repair.³²

In conclusion, the present study demonstrates that surgical trauma suppresses T_H1 cytokine responses without increasing T_H2 cytokine production. These results provide further information that may direct future treatments based on the T_H1 and T_H2 paradigm focused on decreasing the risk of postoperative infectious complications.

Reprints: Ramon Berguer, MD, Surgical Service, Department of Veterans Affairs Northern California Health Care System, 150 Muir Rd, Mail Code 112, Martinez, CA 94553 (e-mail: berguer.ramon@martinez.va.gov).