The basis of cancer therapy is to identify/develop interventions that: a) selectively kill tumor cells while sparing non-malignant cells; b) prevent development of tumor-resistance; and c) act systemically to prevent relapse. Theoretically, immunotherapy of cancer would achieve all these aims. Selective killing of tumor cells has been demonstrated by a wide range of immune cells ranging from conventional CD8 T cells, gamma delta T cells, natural killer (NK) cells, natural killer T (NKT) cells and in some studies neutrophils. Other components of the immune system expressing tumor selectivity include complement factors and natural antibodies of the IgM isotype. Tumor resistance to immunotherapy, differs from resistance to chemotherapy, where expression of multiple drug resistance proteins actively pumps out cancer-toxic substances. One mechanism is downregulation of human leukocyte antigen (HLA) by tumor cells in response to T cell killing. The immune system conceptually overcomes this by the NK subset which preferentially kills cells with downregulated HLA. The other mechanism of tumor resistance from immunological attack is by mutating antigens that are being recognized. Given the promiscuous binding ability of T cell and B cell receptors to bind antigens, the cells could conceptually “mutate with the tumor” in order to recognize and kill cells with variant antigens. Immunological destruction of neoplasia is believed to occur at a systemic level and thus arises the possibility to inhibit metastasis and tumor recurrence, in part through induction of immunological memory.
The history of tumor immunotherapy begins with the work of William Coley who induced a systemic inflammatory/immune activation through administration of killed S. pyogenes and Serratia marcescens bacteria in patients with soft tissue sarcoma [
1]. The advent of molecular biology allowed for assessment of molecular signals associated with systemic immune activation. The cytokine tumor necrosis factor (TNF)-alpha was one of the molecular signals associated with anticancer efficacy of innate immune activators such as the Coley vaccine [
2]. Studies have demonstrated that TNF-alpha has the ability to induce profound death of cancer cells in vitro and in vivo in animal models, however human studies demonstrated unacceptable levels of toxicity [
3‐
5]. IL-2 was the next cytokine associated with immune activation that was tested. Originally termed T Cell Growth Factor (TCGF) [
6], IL-2 was demonstrated in early studies to endow human lymphocytes with ability to selectively kill tumor but not healthy control cells [
7]. Subsequent studies have demonstrated that cytotoxic activity was mediated through T cell and natural killer (NK) cells, whose activation requires stimulation of the IL-2 receptor [
8,
9], which can be accomplished in vivo with high doses of IL-2 [
10‐
12]. Animal studies suggested that IL-2 has a short half-life of approximately 2 minutes after intravenous injection [
13,
14], and human half life was reported to be approximately an hour [
15]. Thus it was apparent that clinical use of IL-2 would be requiring repeated administration at high doses. Despite this pitfall, preclinical studies demonstrated highly potent anti-tumor effect. In 1985 Steven Rosenberg reported regression of established pulmonary metastasis, as well as various subcutaneous tumors by administration of IL-2 [
16]. These data were highly promising due to the fact that tumor killing could be achieved systemically, and by activation of specific immune cells that could be identified in vivo as interacting with and inducing death of the tumor.
Early studies of IL-2 demonstrated impressive results in a subset of melanoma and renal cell cancer patients. Development of systemic autoimmunity to melanocytes, such as the occurence of vitiligo during the treatment with systemic IL-2 was found to be predictive of response. These studies were expanded and eventually IL-2 received approval as the first recombinant immunotherapeutic drug by the FDA. There appears to be a dose response with IL-2 in that the doses that seem to be most effective are also associated with significant toxicity. The most significant cause of toxicity is vascular leak syndrome (VSL), manifested as fluid loss into the interstitial space, which is a result of increase vessel permeability. Additional effects include thrombocytopenia, elevated hepatic serum transaminases, hepatocyte necrosis, hypoalbuminemia, tissue and peripheral eosinophilia, and prerenal azotemia [
17].
Vascular leak syndrome
Vascular leak syndrome (VLS) is considered to be the major dose-limiting adverse effect of IL-2 administration. In a meta-analysis of studies performed in metastatic renal cell carcinoma patients, objective responses were observed in 23% of patients, the majority of which lasted more than 10 years [
18]. Unfortunately, 65% of patients have to interrupt or stop therapy because of VLS [
19,
20]. This syndrome is characterized by an increased extracellular fluid extravasation, hypotension, ascites, pulmonary edema, and hydrothorax [
21,
22], and clinically resembles the systemic inflammatory response syndrome (SIRS). Dermatological manifestations include erythrematous eruptions and mild papillary edema associated with burning and pruritus of the skin [
23,
24]. Severe forms of VLS are associated with pulmonary or cardiac failure with approximately 1% of treated patients having lethal outcome. Typically the symptoms of VLS are treated by vasopressor therapy and judicious fluid replacement, such as with colloid solutions for their osmotic effects. Patients may also be treated with theophylline and terbutaline, for which clinical experience suggests a possible reduction of the severity and frequency of acute episodes [
25].
Oxidative stress
Numerous studies have demonstrated that oxidative stress modifies endothelial cells in a manner that preferentially activates the complement cascade [
62]. The involvement of the mannose-binding lectin and the lectin complement pathway (LCP) in promoting complement activation by endothelial cells post oxidative stress was shown in studies using hypoxic (24 hours; 1% O(2))/reoxygenated (3 hours; 21% O(2)) human endothelial cells. Using iC3b deposition as a marker of complement activation, it was shown that N-acetyl-D-glucosamine or D-mannose, but not L-mannose, blocked activation, suggesting that oxidative stress upregulates the mannose dependent pathway. This was also demonstrated using mannose binding lectin deficient serum, as well as antibodies to mannose binding lectin. Furthermore C3 deposition was found in ischemic areas in rats that experienced cardiac ischemia reperfusion injury, a known inducer of oxidative stress [
63].
Vitamin C Depletion after IL-2 therapy
The initial stimulus for writing this review was inspired by the observations made by two of the authors (DTA, CAD) of a scurvy-like condition in a renal cell carcinoma patient treated with IL-2. The patient presented with acute signs and symptoms of scurvy (perifollicular petechiae, erythema, gingivitis and bleeding). Serum ascorbate levels were significantly reduced to almost undetectable levels during the treatment with IL-2 [
64]. Although the role of ascorbic acid (AA) hyper-supplementation in stimulation of immunity in healthy subjects is controversial, it is well established that AA deficiency is associated with impaired cell mediated immunity. This has been demonstrated in numerous studies showing that deficiency of this vitamin suppresses T cytotoxic responses, delayed type hypersensitivity, and bacterial clearance [
65]. Additionally, it is well-known that NK activity, which mediates IL-2 anti-tumor activity, is suppressed during conditions of AA deficiency [
66]. Thus it may be that while IL-2 therapy on the one hand is stimulating T and NK function, the systemic inflammatory syndrome-like effects of this treatment may actually be suppressed by induction of a negative feedback loop. This negative feedback loop with IL-2 therapy was successfully overcome by work using low dose histamine to inhibit IL-2 mediated immune suppression, which led to the “drug” Ceplene (histamine dichloride) receiving approval as an IL-2 adjuvant for treatment of AML [
67].
The concept of AA deficiency subsequent to IL-2 therapy was reported previously by another group. Marcus et al evaluated 11 advanced cancer patients suffering from melanoma, renal cell carcinoma and colon cancer being on a 3 phase immunotherapeutic program consisting of: a) 5 days of i.v. high-dose (10(5) units/kg every 8 h) interleukin 2, (b) 6 1/2 days of rest plus leukapheresis; and (c) 4 days of high-dose interleukin 2 plus three infusions of autologous lymphokine-activated killer cells. Mean plasma ascorbic acid levels were normal (0.64 +/- 0.25 mg/dl) before therapy. Mean levels dropped by 80% after the first phase of treatment with high-dose interleukin 2 alone (0.13 +/- 0.08 mg/dl). Subsequently plasma ascorbic acid levels remained severely depleted (0.08 to 0.13 mg/dl) throughout the remainder of the treatment, becoming undetectable (less than 0.05 mg/dl) in eight of 11 patients during this time. Importantly, blood pantothenate and plasma vitamin E remained within normal limits in all 11 patients throughout the phases of therapy, suggesting the hypovitaminosis was specific for AA. Strikingly, responders (n = 3) differed from nonresponders (n = 8) in that plasma ascorbate levels in the former recovered to at least 0.1 mg/dl (frank clinical scurvy) during Phases 2 and 3, whereas levels in the latter fell below this level [
68]. Similar results were reported in another study by the same group examining an additional 15 patients [
69]. The hypothesis that prognosis was related to AA levels is intriguing because of the possibility of higher immune response in these patients, however this has not been tested.
AA protects the endothelium from inflammation
The main cause of VSL is increased permeability of the endothelium. Regardless of if the initiating cause is T cell activation, complement, and/or oxidative stress, the effector mechanism of VLS is alteration of endothelial cell function. In SIRS the endothelium is also the main effector causing lethality. Methods by which these cells are altered by both SIRS and IL-2 include: a) endothelial cell apoptosis, b )upregulation of adhesion molecules, and c) increased procoagulant state [
70].
It was shown that in vitro administration of AA led to reduction of TNF-alpha induced endothelial cell apoptosis [
71]. The effect was mediated in part through suppression of the mitochondria-initiated apoptotic pathway as evidenced by reduced caspase-9 activation and cytochrome c release. Another study examined 34 patients with NYHA class III and IV heart failure who received AA or placebo treatment. AA treatment (2.5 g administered intravenously and 3 days of 4 g per day oral AA) resulted in reduction in circulating apoptotic endothelial cells in the treated but this was not observed in the placebo control group [
71]. Various mechanisms for inhibition of endothelial cell apoptosis by AA have been proposed including upregulation of the anti-apoptotic protein bcl-2 [
72] and the Rb protein, suppression of p53 [
73], and increasing numbers of newly formed endothelial progenitor cells [
74].
AA has been demonstrated to reduce endothelial cell expression of the adhesion molecule ICAM-1 in response to TNF-alpha in vitro in human umbilical vein endothelial (HUVEC) cells (HUVEC) [
75]. By reducing adhesion molecule expression, AA suppresses systemic neutrophil extravasation during sepsis, especially in the lung [
76]. Other endothelial effects of AA include suppression of tissue factor upregulation in response to inflammatory stimuli [
77], an effect expected to prevent the hypercoaguable state. Furthermore, ascorbate supplementation has been directly implicated in suppressing endothelial permeability in the face of inflammatory stimuli [
78‐
80], which would hypothetically reduce vascular leakage. Given the importance of NF-kappa B signaling in coordinating endothelial inflammatory changes [
81‐
83], it is important to note that AA at pharmacologically attainable concentrations has been demonstrated to specifically inhibit this transcription factor in endothelial cells [
84]. Several pathways of inhibition have been identified including reduction of i-kappa B phosphorylation and subsequent degradation [
85], and suppression of activation of the upstream p38 MAPK pathway [
86]. In vivo data in support of eventual use in humans has been reported showing that administration of 1 g per day AA in hypercholesterolemic pigs results in suppression of endothelial NF-kappa B activity, as well as increased eNOS, NO, and endothelial function [
87]. In another porcine study, renal stenosis was combined with a high cholesterol diet to mimic renovascular disease. AA administered i.v. resulted in suppression of NF-kappa B activation in the endothelium, an effect associated with improved vascular function [
88].
AA Suppresses Systemic Inflammation
The possibility that IL-2 therapy induces a state of systemic inflammation similar to SIRS has been discussed previously. One of the fundamental questions is whether AA actually has beneficial effects on the process of systemic inflammation. A mouse study demonstrated that after challenge with the bacteria Klebsiella pneumonia to induce a sepsis-like state, a 3-fold higher mortality is observed in ascorbate-deficient animals compared to controls [
89]. Another study hyper-supplemented animals with AA by administration of 10 mg/kg AA intravenously before induction of sepsis. IV AA treated animals had a 50% survival while only 19% of control animals survived [
90]. Other studies demonstrated that hyper-supplementation with AA resulted in better outcomes in sepsis-associated hypoglycemia [
91], microcirculatory abnormalities [
92], and blunted endothelial responsiveness [
93‐
95] in animal models.
Randomized clinical trials have been performed in septic patients using AA and vitamin E, which demonstrated superior outcomes, as well as reduction in parameters of oxidative stress [
96,
97]. To date we know of one study in the recent history that assessed AA alone in patients with systemic inflammation. The investigators examined burn patients with > 30% of their total body surface area affected. Patients were administered intravenous AA i.v. (66 mg/kg/hr for 24 hours, n = 19) or received only standard care (controls, n = 18). AA treatment resulted in statistically significant reductions in 24 hr total fluid infusion volume, and fluid retention (indicative of vascular leakage). Perhaps most striking was the decrease in the need for mechanical ventilation: the treated group required an average of average of 12.1 ± 8.8 days, while the control group required 21.3 ± 15.6 days [
98]. Given that numerous inflammatory markers associated with VLS are also found in SIRS and severe burn patients, the possibility is presented that AA may exert some beneficial effects on IL-2 therapy, both from the reduction of toxicity perspective, as well as from the stimulation of efficacy.