Immune responses in patients with brain tumors. Factors such as anti-convulsants that may contribute to impaired cell-mediated immunity

Prophylactic antibiotics and anticonvulsants in neurosurgery

The prophylactic administration of antibiotics to prevent infection and the prophylactic administration of anticonvulsants to prevent first seizure episodes are common practice in neurosurgery. If prophylactic medication therapy is not indicated, the patient not only incurs the discomfort and the inconvenience resulting from drug treatment but is also unnecessarily exposed to adverse drug reactions, and incurs extra costs. The main situations in which prophylactic anticonvulsants and antibiotics are used are described and those situations we found controversial in the literature and lack further investigation are identified: anticonvulsants for preventing seizures in patients with chronic subdural hematomas, antiepileptic drugs for preventing seizures in those suffering from brain tumors, antibiotic prophylaxis for preventing meningitis in patients with basilar skull fractures, and antibiotic prophylaxis for the surgical introduction of intracranial ventricular shunts.In the following we present systematic reviews of the literature in accordance with the standard protocol of The Cochrane Collaboration to evaluate the effectiveness of the use of these prophylactic medications in the situations mentioned. Our goal was to efficiently integrate valid information and provide a basis for rational decision-making.

Dexamethasone-induced abolition of the inflammatory response in an experimental glioma model: a flow cytometry study

OBJECT

Commonly used for management of cerebral edema in patients with brain tumors, steroid medications also have immunosuppressive functions. To characterize the effects of steroids on the central nervous system's response to tumors more clearly, flow cytometry was used to quantify the extent of inflammatory cell infiltration in an immunogenic rat glioma model.

METHODS

Freshly prepared 11-day-old intracranial C6 tumors that had been excised from dexamethasone-treated and untreated rats were labeled ex vivo with monoclonal antibodies against CD 11b/c, CD45, and CD8a antigens. The extent of microglia (CD11b/c-highly positive, CD45-slightly positive cell), macrophage (CD11b/c-highly positive, CD45-highly positive cell), lymphocyte (CD11b/c-negative, CD45-highly positive cell), and cytotoxic T-cell (CD8a-positive cell) infiltration into each rat's tumor, tumor periphery, and contralateral tumor-free hemisphere was analyzed using flow cytometry. Microglia and lymphocytes constituted a significant component of infiltrating cells in this model, comprising 23 +/- 3% and 33 +/- 5% of viable cells, respectively. Macrophages, on the other hand, accounted for only 9 +/- 1% of infiltrating cells. Treatment of rats with a 7-day course of low-dose dexamethasone (0.1 mg/kg/day) resulted in a greater than 50% inhibition of microglia (p = 0.03) and lymphocyte (p = 0.001) infiltration into tumors. Increasing the dexamethasone dose to 1 mg/kg/day further abolished lymphocyte infiltration (89% inhibition, p = 0.001) but had no additional inhibitory effect on microglia invasion. Macrophage infiltration of tumors was not inhibited at the dexamethasone doses used in this study (p = 0.42).

CONCLUSIONS

Flow cytometry is a valuable technique for characterizing tumor-associated inflammatory cells in gliomas. Even at low doses, dexamethasone was found to inhibit significantly the infiltration of brain tumors by lymphocytes and microglia. These findings should be considered when experimental immunotherapeutic strategies are evaluated for clinical application.

Anticonvulsant-induced suppression of IFN-gamma production by lymphocytes obtained from cervical lymph nodes in glioma-bearing mice

It is well known that phenytoin can cause impairment of cellular immunity. The authors investigated the potential role of other anticonvulsant drugs in the development of antitumor immunity in murine malignant glioma models. The survival rate was determined in murine glioma models using syngeneic 203 glioma cells following treatment with four anticonvulsants, which are most commonly administered to glioma patients, i.e., phenytoin, phenobarbital, valproate and zonisamide. In a second set of experiments, we further examined the effect of these drugs on interferon-gamma (IFN-gamma) secretion by lymphocytes prepared from cervical lymph nodes (CLN) in the same models. The IFN-gamma production of CLN lymphocytes as measured by ELISA method was markedly impaired in the early stage of tumor-bearing mice treated with phenytoin or zonisamide, and the median survival time (MST) of controls and of mice treated with either phenytoin or zonisamide was 13, 10 and 11 days, respectively, which was not a statistically significant difference. Phenobarbital and valproate did not affect either IFN-gamma production or their survival rate. In addition, immunohistochemistry showed a reduction in tumor-infiltrating lymphocytes containing CD4 and CD8 antigens in the mice treated with phenytoin and zonisamide. Two anticonvulsants, phenytoin and zonisamide, showed a significant inhibitory effect on IFN-gamma production by CLN lymphocytes in murine glioma models, although there was no statistically significant difference in MST between controls and the anticonvulsant-treated mice. These drugs might have some detrimental influence on the prognosis of brain tumor patients when combined with the latent immune dysfunction accompanying the tumor-bearing state.

Chronic staphylococcal scalded skin syndrome

Staphylococcal scalded skin syndrome (SSSS), not previously recorded as a chronic disease, persisted for 2 years in a 50-year-old woman with epilepsy and cerebellar ataxia. Lesions initially suggestive of erythema multiforme and toxic epidermal necrolysis evolved over 2 years into those typical for SSSS, with extensive erosions and subcorneal blisters, showing an epidermal split at the granular cell layer. Exfoliatin A-producing phage I-III Staphylococcus aureus, previously linked only to acute mild adult cases of SSSS, was cultured from purulent discharge in the patient's eyes, ears and open skin lesions. The roles of epilepsy and antiepileptic medications are discussed as possible predisposing factors.

Enhanced interleukin-1 beta release and longevity of glioma-associated peripheral blood monocytes in vitro

Interleukin-1 (IL-1) plays a controversial role in the immune response. Besides its activation of immune cells and juvenile central nervous system cells, monocyte-derived IL-1 may be able to stimulate the malignant transformation and proliferation of glial brain tumor cells expressing IL-1 receptors. The aim of this study was to determine the growth pattern and the IL-1 beta release of long-term cultured peripheral blood monocytes of glioma patients. At 6- to 7-day intervals, the vital monocytes, characterized by CD14 immunophenotyping, were counted. By the use of a specific IL-1 beta enzyme-linked immunosorbent assay, the IL-1 beta content of monocyte culture supernatants derived from 13 subjects with glioma and from 12 controls were compared at Days 7, 21, and 100 of culture. Cell clusters of monocytes derived from glioblastoma patients survived more than 250 days in culture, whereas control monocytes survived only up to 114 days. The IL-1 beta release of glioma-associated peripheral blood monocyte cultures was about 50 times higher as compared with control monocyte cultures. Dexamethasone treatment at the time of blood sampling and recurrences of the gliomas did not influence the increase in the IL-1 beta expression of glioma monocytes. It seemed that at least subsets of glioma-associated blood monocytes, although they had been removed from the circulation, remained activated for a long period of time. We conclude that increased IL-1 beta production of glioma-associated peripheral blood monocytes and their longevity in vitro may be features of aberrant immune cell subsets. In future studies, the exact phenotyping of monocyte subsets will be mandatory.

A correlation between serum immunosuppressive acidic protein and altered immunocompetence in patients with brain tumours

The present investigation was conducted to document a correlation between the serum levels of immunosuppressive acidic protein (IAP) and depressed lymphocyte responsiveness to mitogens in vitro in patients with intracranial tumours, and to delineate the possible roles of IAP upon immunocompetence in these patients. It was thought that high concentrations of IAP present in the serum of brain-tumour patients may play a significant role in the immunosuppression seen in this patient population. The effect of IAP upon mitogen-stimulated lymphocyte function was evaluated by tritiated (3H)-thymidine incorporation. Lymphocytes from both 30 patients with intracranial tumours and 30 normal individuals were incubated for 90 hours in culture medium in the presence of three mitogens: phytohaemagglutinin (PHA), concanavalin A (Con A), and pokeweed mitogen (PWM). Lymphocytes obtained from patients with brain tumours and cultured in autologous serum displayed a significant depression of 3H-thymidine incorporation, as was observed in previous studies. In addition, a significant suppression of mitogen-induced activation of the normal lymphocytes was demonstrated in the presence of allogeneic patients' serum and the percentage of inhibition was found significantly proportional to the IAP concentrations. Furthermore it was also demonstrated that increased levels of serum IAP could significantly correlate with two in vivo aspects of impaired cellular immunity: the decreased lymphocyte counts in the peripheral blood and diminished cutaneous delayed hypersensitivity reactions measured by purified protein derivative (PPD) skin test reactivity. On the other hand, an attempt was also made to investigate changes in humoral immunity and immunoglobulin concentrations were observed not to correspond to the serum IAP levels. These studies suggest a possible connection between serum IAP levels and altered cellular immune competence in brain-tumour patients.

Growth factors in the tumorigenicity of a brain tumor cell line

We studied a virally induced canine gliosarcoma brain tumor cell line capable of producing brain tumors in vivo following intracerebral inoculation of the cells into the brains of adult mongrel dogs. Cloned populations from this cell line differed in their ability to produce brain tumors in vivo. In vitro transforming growth factor-beta secretion by each clone correlated with in vivo tumorigenicity. Mitogenic activity like that produced with platelet-derived growth factor was also secreted by each clone and correlated with in vivo tumorigenicity. Epidermal growth factor and transforming growth factor-alpha production was not detected in any of the clones of canine gliosarcoma cells. Although only some of the brain tumor cell clones were able to produce tumors in immunologically normal dogs, brain tumors were produced in all dogs pretreated with cyclosporin and in all "nude" mice, suggesting that tumor immunology plays an important role in this brain tumor model. In vitro cloning efficiency of each brain tumor cell clone also correlated with in vivo tumorigenicity. We postulate that transforming growth factor-beta may both stimulate brain tumor cell growth and inhibit host antitumor immune surveillance.

Cell-mediated cytotoxicity in glioma-bearing patients: differential responses of peripheral blood mononuclear cells to stimulation with interleukin-2 and microbial antigen

Peripheral blood mononuclear cells (PBMC) of malignant glioma-bearing patients were stimulated in vitro with Interleukin-2 (IL-2) or a glucomannan-protein antigen of Candida albicans (GMP) then assayed for proliferation, production of IFN-gamma, and generation of cytotoxic effectors against either K562 tumor cell line or freshly-cultured allogenic glioma cells. PBMC of healthy, age and sex-matched subjects were the controls. PBMC of glioma-bearing patients did not differ, as a whole, from PBMC of healthy donors in IL-2 or GMP-induced proliferation. However, they showed a lesser ability to produce IFN as well as a substantial inability to generate cytotoxic effectors following GMP stimulation. PBMC of glioma patients were fully responsive to IL-2 in cytotoxicity generation, as were the PBMC from normal subjects. The results suggest that glioma patients may have a defective antigen-mediated activation of natural cytotoxic effectors. This hyporesponsiveness is not accompanied by depressed lymphoproliferation and does not apparently involve a reduced response to IL-2.

Comparison of in vitro glioma cell cytotoxicity of LAK cells from glioma patients and healthy subjects

Peripheral blood mononuclear cells from 11 glioma patients and 11 healthy control subjects were cultured in medium containing recombinant interleukin-2 for a period of 5 days. The cytotoxicity of these lymphokine-activated killer (LAK) cells was tested on chromium-51-labeled freshly prepared allogeneic glioblastoma cells, and on the cell lines K562 (natural killer cell (NK)-sensitive) and Daudi (NK-resistant). Peripheral blood mononuclear cells from all subjects showed high levels of cytotoxicity against these targets. There was no significant difference between the patients and the control group when LAK cytotoxicity was compared. Thus, although glioma patients are known to have depressed immunological reactivity, the cytotoxic capacity of LAK cells derived from glioma patients is similar to that of LAK cells from healthy control subjects. However, the glioma patients had significantly reduced numbers of mononuclear cells in their peripheral blood, possibly due to steroid treatment. Therefore, the volume of blood required to generate the same number of LAK cells was approximately three times larger from the glioma patients than from control subjects.

Imbalances in T cell subpopulations in human gliomas

The quantitation of cells bearing CD3, CD4, CD8, and B cell phenotypic markers, as well as an estimation of serum immunoglobulin (Ig)G, IgA, and IgM, was carried out in a group of 39 glioma patients with different grades of malignancy. The findings were compared with those obtained from 21 normal healthy control subjects. The analysis revealed a significant decrease both in the absolute numbers and in the percentages of circulating CD3+ (p less than 0.001) and CD4+ (p less than 0.001) cells, while the CD8+ and Pan B+ cells remained within the normal range irrespective of the type and grade of tumor. The CD4+:CD8+ ratio was significantly decreased in all categories of patients. The CD4 lymphopenia was also evident in 10 patients who had no history of previous immunosuppressive drug therapy (steroids and anticonvulsants) until the commencement of the study. The Ig levels were within the normal range in patients with malignant astrocytoma and glioblastoma multiforme, whereas a three- and fourfold increase in the IgM level was observed in patients with astrocytoma. It is suggested that T cell lymphopenia in glioma patients could mainly be due to a selective depletion of CD4+ cells and that it occurs principally as a reaction to the tumor.

The response of peripheral blood mononuclear cells of glioma-bearing patients to stimulation with microbial antigen and IL-2: proliferation and IFN-gamma production

Peripheral blood mononuclear cells (PBMC) of malignant glioma-bearing patients were stimulated in vitro with a glucomannan-protein antigen of Candida albicans (GMP) or Interleukin-2 (IL-2), then assayed for proliferation and production of Interferon gamma. PBMC of healthy, age and sex matched subjects were the controls. PBMC from glioma-bearing patients did not differ, as a whole, from PBMC of healthy donors in IL-2 or GMP-induced proliferation. However, they showed a definitely lesser ability to produce IFN. The results are discussed in the framework of the impairment of immune responses known to affect glioma-bearing patients.

Effect of phenytoin on antibody production: use of a murine model

An animal model was used to study the effects of phenytoin on immune function. Inbred NFS mice, injected with phenytoin, 10-40 mg/kg, and immunized with bovine serum albumin (BSA), demonstrated a dose-dependent decrease in production of IgG antibodies to BSA. Phenytoin at the dosages used in this study did not appear to be overtly toxic to immune organs and cells as judged by normal weights of thymus and spleen and normal WBC counts of mice injected with phenytoin. In addition, mice receiving phenytoin had normal body weights, liver weights, and hematocrits. The usefulness of an animal model to study immunomodulatory effects of antiepileptic drugs is discussed.

Blood lymphocyte subsets in patients with primary intracranial tumours. Correlation to histological tumour type and anatomical site

The blood lymphocyte population of 118 patients with primary intracranial tumours and healthy volunteers was examined with respect to its size and cellular composition using various rosette tests. The patients had not undergone any surgical intervention or received any treatment with ionizing irradiation or cytotoxic drugs. However, some of them were treated with corticosteroids. It was observed that non-steroid treated patients with oligodendrogliomas, but not patients with other histological types of tumours, had a significantly reduced proportion of "active" T-lymphocytes forming rosettes with sheep erythrocytes (a type of T-lymphocyte which is activated by foetal calf serum). These patients as well as those with astrocytomas, malignant gliomas (anaplastic astrocytomas and glioblastomas) or miscellaneous tumours (mainly meningiomas) had normal proportions of lymphocytes with receptors for the Fc-part of IgG or C'3 and cells forming rosettes with sheep erythrocytes under more conventional conditions. Patients who were treated with corticosteroids had an increased frequency of lymphocytes with the above Fc-receptor. An association between site of the lesions and cellular composition of the blood lymphocyte population was not detected. The results give further support for the view that the immunological system may be changed in patients with oligodendrogliomas.

Immunosuppression by phenytoin: implication for altered immune competence in brain-tumor patients

This investigation was conducted to examine the immunosuppressive potential of phenytoin in vivo and to document a correlation between phenytoin therapy and depressed lymphocyte responsiveness to mitogens. It was thought that phenytoin, the most widely used anticonvulsant agent, may play some role in the immunosuppression seen in brain-tumor patients. The effect of phenytoin on mitogen-stimulated lymphocyte function was evaluated by tritiated (3H)-thymidine incorporation and lymphocyte nuclear size distribution. Lymphocytes from either phenytoin-treated or normal rabbits were incubated for 90 hours in culture medium in the presence of three mitogens: phytohemagglutinin (PHA), concanavalin A (Con A), and pokeweed mitogen (PWM). Significant suppression of mitogen-induced activation of the lymphocytes from treated animals was demonstrated. The present studies suggest a possible connection between phenytoin therapy and altered immune competence in brain-tumor patients.

Presence of immunosuppressive factors in brain-tumor cyst fluid

The present investigation was conducted to examine the effects upon normal lymphocyte activation of the addition of brain-tumor cyst fluid or cerebrospinal fluid (CSF) to an in vitro culture system. It was believed that factors present in these fluids may play a role in local and systemic immunosuppression in brain-tumor patients. The authors evaluated the effect of the cyst fluid, CSF, and serum from eight patients with brain tumors (one astrocytoma, five glioblastoma multiforme, one medulloblastoma, and one microglioma) on mitogen-stimulated lymphocyte function as measured by 3H-thymidine incorporation, cell size distribution, and cellular proliferation. Lymphocytes from normal volunteers were incubated for 90 hours in culture medium with 10% pooled human serum, mitogens (phytohemagglutinin, concanavalin A, and pokeweed mitogen), and a range of volumes of cyst fluid or CSF from the tumor patients. Significant suppression of mitogen-induced activation of the lymphocytes was demonstrated in the presence of cyst fluid from five out of six patients, but the mitogen response was only minimally affected by the addition of the CSF from these patients to the culture medium. The suppression of lymphocyte activation by cyst fluid was directly proportional in several cases to the concentration of the cyst fluid. The total protein, albumin, and immunoglobulin (Ig) concentrations in the cyst fluids were observed to roughly correspond to the serum levels and were much higher than in CSF. These studies suggest that brain-tumor cells locally produce lymphocyte-suppressive factors which may then be released into blood. Preliminary characterization of the suppressive factor(s) responsible for the inhibition of lymphocyte function indicates that they are non-dialyzable and do not appear to be an IgG.