Clinical and immunological allergy assessment in cancer patients

Cancer is associated with immunodeficiency, while allergies result from immune system hyperactivity mediated by cytokines and immunoglobulins. The purpose of this study was to determine the relationship between immune environment of specific cancers and allergies, emphasizing cytokines related to Th1 and Th2 responses associated with IgE. 80 adults were distributed into two groups: control (n = 20) and cancer (n = 60), distributed in three subgroups (n = 20), head and neck, stomach, and prostate cancers. This study compared Th1 (IL-2) and Th2 (IL-4) parameters, anti-inflammatory, pro-inflammatory, or regulatory profile regarding both IgE levels and reported allergies, by means of clinical manifestations and IgE, IL-1β, IL-2, IL-4, IL-17, and TGF-β serum concentration. Clinically allergies were observed in 50% of the control group and in 20% of the cancer group (p = 0.009). IL-2 cytokine and TGF-β concentrations were higher in the patients with cancer as compared to the control (p < 0.005). However, there were IL-4, IL-17, and IL-1β decreases in the patients with cancer (p < 0.05). No correlation was observed between the cytokines studied and IgE and clinically proven allergies in both investigated groups. There was an inverse association between cancer and clinical allergy manifestations. In head and neck, stomach, and prostate cancers, an immunosuppressive serum tumor environment was predominant. There was no difference in cytokines related to Th1 and Th2 parameters in relation to IgE. No correlation was found between clinically proved allergies and immunity markers related to the same allergens.


Scientific Reports
| (2021) 11:18110 | https://doi.org/10.1038/s41598-021-97200-y www.nature.com/scientificreports/ and cancer 27,28 . The transforming growth factor-β (TGF-β) plays a two-phase role in cancer development and advances: in normal cells, TGF-β suppresses tumor genesis. However, once the tumor process has already begun, TGF-β creates an immunosuppressive microenvironment by inhibiting IL-2, IL-12, and IFN-γ. Therefore, local immune tolerance is created, and the TGF-β serum concentration is increased, which is frequent in cancer patients [29][30][31][32] . The allergy process is assessed by clinical assessment, skin tests, and total or specific IgE serum levels. However, no direct relation between specific reported allergies nor skin sensitivity to the same allergen has been observed, understanding that a clinical allergy may be accompanied by a negative skin test for the same sensitivity agent 33 . Positive skin tests are observed in people who do not exhibit clinical allergies. In the presence of cancer, a lower rate of reported allergies was observed for the general population, even with a higher incidence of positive skin tests for allergens 33 . The IgE serum measurement (type I hypersensitivity reaction marker) has been used as an objective diagnostic method for allergies. The inverse association between IgE concentration and cancer may be specific to the allergy and tumor types 6,20,34,35 .
The mechanism proposed to explain the association between cancer and allergy is that in allergy a heightened immune response is prevalent, whereas in cancer a hyporesponsive immune response is observed. Thus, in allergic people there would be an increase in production of cytokines related to the Th2 response and increasing the production of IgE against the specific tumor antigen, capable of activating effector cells and destroying tumors [36][37][38][39] . However, few studies investigate the measurement of cytokine and other biomarkers in this situation 36,37 . Almost all trials involving cytokines are conducted in animal models 19 , and only two studies with human subjects have been identified 38,39 .
The purpose of this study, which is part of a line of research 33,40,41 , was to determine the relationship between immune environment of specific cancers and allergies, emphasizing cytokines related to Th1 and Th2 responses associated with IgE.

Results
Demographic data are described in Table 1. There was a predominance of male subjects in the cancer patient group, and the prevalence of alcoholism and smoke was higher in the cancer group. The average age of the participants was 63 (± 10) years and there was no difference between the groups (p > 0.05). Cancer staging was I-3%, II-24%, III-38%, and IV-35%. There was an inverse relation in the comparison between the control group and the prostate adenocarcinoma group (OR 0.17, 95% CI 0.04-0.79, p = 0.024). However, no difference was observed in the comparison with the gastric adenocarcinoma group (OR 0.25, 95% CI 0.06-1.02, p = 0.053), or with the head and neck squamous cell carcinoma (HNSCC) group (OR 0.33, 95% CI 0.09-1.27, p = 0.108). The most common type of allergy was rhinitis, which was observed in 15% of the cancer patients and in 35% of the control individuals (p = 0.005) ( Table 1).
The agreement between reported allergies and positive IgE (considered to be an IgE concentration ≥ 25 UI/ml) was weak in all groups: κ = 0.000 in the control group, κ = 0.002 in the cancer patient group, namely κ = −0.037 for prostate adenocarcinoma, κ = 0.237 for gastric adenocarcinoma, and κ = −0.250 for HNSCC.
There was no difference in the values of cytokine medians between the two groups, based on IgE concentrations or on the classification of IgE in the normal, borderline, or high categories (p > 0.05) ( Table 3). There was a weak correlation between most cytokines and IgE levels in both groups (Spearman's correlation factors < 0.4 for most groups) 45 . A moderate correlation was identified only in the prostate adenocarcinoma group for TGF-β (r = 0.54, p = 0.01) and IL-17 (r = 0.42, p = 0.06) as well as for gastric adenocarcinoma (r = 0.40, p = 0.07) for IL-2 ( Fig. 1). There was no correlation between IgE concentrations and staging in the general cancer group (r = 0.25, p = 0.60), prostate cancer (r = 0.12, p = 0.62), HNSCC (r = 0.07, p = 0.78), and stomach cancer (r = 0.22, p = 0.37).
To determine whether the presence of reported allergies is associated with the predominance of some immunity markers, cytokines (Th1 (IL-2), Th2 (IL-4), Th17 (IL-17), IL-1β, TGF-β) and IgE were compared between both groups. Table 4 shows there was no difference among these variables in this study.

Discussion
The relation between cancer and allergies has been studied since the 1950s 46 , and has yielded conflicting results, due to the diversity of study methods and to risk factors for cancer or allergies, such as age, use of medication and associated diseases, often not controlled 6,38,39 . These variables were observed and carefully adjusted in this study, to avoid interfering in the interpretation of results and to preserve their reliability.
There was an inverse association between cancer and clinically detected allergies, which favors the theory of immune surveillance 12,47 , which is higher in allergic processes and may be associated with the damage on malignant cells 1,5 . In this study, as well as in other investigations conducted in the same line of research, there was a lower prevalence of allergies in cancer patients, especially in the presence of prostate adenocarcinoma, in comparison with the control group 33,40 . There was also an inverse association between reported allergies and the presence of HNSCC, as well as of gastric adenocarcinoma 34,37,48,49 . The lack of significance in this comparison in the present study, which was revealed only as a trend, is probably due to the small number of cases, although we have studied a larger number of individuals than has been described in the literature in similar studies 38,39 .  www.nature.com/scientificreports/ A pivotal limitation of this work was the low number of participants per group due to the specificity of cancers, which were investigated. The rigorous statistical analysis of all parameters leads this study suitable to drive conclusions concerning the relationship between allergies and cancer. Other types of cancer could be included in this study, which belong to future surveys of the same line of research. These upcoming studies will expand the current data with more participants.
Allergy is a complex disorder that involves immune factors that may or may not depend on IgE, and few studies are available that evaluate the IgE serum dosage in cancer patients 35,[50][51][52] . Although IgE is useful in detecting allergies, this immunoglobulin is not increased in all types of allergic reactions, which may explain  www.nature.com/scientificreports/ the low concordance between reported allergies and positive IgE observed in this work, likewise as was reported in another study 44 . When the inverse association between cancer and allergies is detected, immune surveillance and immunoprophylaxis theories are mentioned as an explanation. Immune surveillance theory suggests that reactions to allergens increase the immune response, including the production of IgE and the activation of effector immune cells (monocytes, macrophages, natural killer cells, eosinophils, mastocytes, and basophils), capable of detecting and destroying cancer cells. Activation of the Th2 response observed in allergic conditions promote the synthesis of related cytokines, such as IL-4 and IL-3, which favors the synthesis of IgE against specific tumor antigens, capable of activating effector cells and destroying tumors 19,36,37 . Immunoprophylaxis theory is related to inhaled  Only two studies evaluated the measurement of cytokines and other biomarkers, connecting cancer and allergies 38,39 . In a study comparing cytokines from HNSCC to the allergic rhinitis group and the control group (with no cancer or allergies), no difference was observed among the concentrations of the studied cytokines (response Th1 (IL-2, IL-12, IFN-γ, TNF-α), Th2 (IL-4, IL5, IL-13), innate immunity (IL-1β, IL-8, IL-17), innate immunity (MCP)-1, macrophage inflammatory protein (MIP)-1β, granulocyte and macrophage colony stimulating factor (G-CSF), granulocyte and macrophage colony stimulating factor (GM-CSF), and immunity related to the T cells (IL-6, IL-7, IL-10) 38 . A different study with breast cancer patients revealed no differences in cytokine concentrations (IL-1β, IL-6, IFN-γ, IL-4) between the cancer and allergy groups 39 .
The fact that groups were not paired by age, which may have interfered in the results, as the immune response changes with age, and that a lower number of patients, if compared to the present work, was involved, are limitations in these two previous studies 38,39 . Results in this study, as well as in those mentioned in the literature, indicated no difference in cytokines related to Th1 and Th2 parameters in the association between cancer and allergic processes related to type I hypersensitivity reactions. This finding contradicts one of the proposed theories in epidemiological study literature about cancer and allergy, in which a difference in cytokines related to Th1 and Th2 parameters had been suggested 19,36,37 .
New studies have increased the knowledge beyond the immune system's Th1/Th2 paradigm and other subsets of cells involved in this process, such as Th3, Th17, Th9, follicular Th, and regulatory T, illustrating that these interrelations are broader. An additional hypothesis, other than the Th1/Th2 paradigm, is based on publications arguing that T CD4+ lymphocytes exhibit plasticity and do not behave as inflexible Th1 or Th2, producing only a   53,54 , despites being flexible in the synthesis of mediators and in the conversion of effector cells Th1, Th2, Th17, or regulatory T [21][22][23]55 . Based on the action of certain cytokines, a subset of T cells may be transformed into different cytokine secretory cells [21][22][23]55,56 . Immune modulating cytokines (TGF-β, IL-10, IL-17, and IL-23) facilitate tumor development, while others (IFN-γ, TNF-α, GM-CSF, IL-2, IL-6, IL-17, and IL-12) promote immune surveillance and delay its growth 19,57 . Studies discussing cancer and allergies must evaluate these literature findings.
To determine whether there is a correlation of other mediators potentially involved in the relation between cancer and allergies, other cytokines were studied. An increase in regulatory cytokine IL-2 and TGF-β was detected, as was the reduction of IL-4, IL-17, and IL-1β, reflecting the cancer patient's immunosuppressive environment and the lower prevalence of allergies. Such findings and characteristic cytokine levels in tumors confirm the specificity of the immune response according to the type of cancer 32 . The moderate correlation between IgE and IL-17 and TGF-β in the prostate adenocarcinoma group suggests the presence of an allergy marker with immunosuppressive response and immune response specificity, according to the tumor type. Studies show that TGF-β does not have a direct effect on the origin of the Th17 cells; however, it may favor its development to suppress the Th1 differentiation 58,59 .
A better understanding of the mechanisms involved in the relationship between cancer and allergies require studies involving a higher number of cytokines and their relation with the molecular messenger (messenger RNA) responsible for synthesizing the cytokines involved in allergies and inflammation 23,60 . Future studies should involve the evaluation, in control environments, of the immune system to exogenous antigens that trigger allergies in cancer patients, in a dynamic study of antigen stimulus and its response in the immune system [61][62][63] .

Conclusions
There was an inverse association between cancer and clinical allergy manifestations. In head and neck, stomach, and prostate cancers, an immunosuppressive serum tumor environment was predominant. There was no difference in cytokines related to Th1 and Th2 parameters in relation to IgE. No correlation was found between clinically proved allergies and immunity markers related to the same allergens. This study included patients with a recent histological diagnosis of cancer and they were only admitted to the research before undergoing any type of treatment. The individuals included were older than 18 years of age, from both genders, who presented an Eastern Cooperative Oncology Group (ECOG) performance score between zero and two 42 , TNM staging from I to IV 43 , and no previous history of other cancer, autoimmune disease or immune treatment, corticosteroid therapy, chemotherapy, or radiation therapy.

Methods
The control group consisted of 20 volunteers, with no cancer and using no medications, selected by age to be paired with the cancer group. The clinical assessment was conducted to identify addictions, alcoholism and smoking, associated diseases, medications, allergy history, allergy type (allergic rhinitis, atopic dermatitis, asthma, food and drug allergies), as well as allergic symptoms after exposure to certain allergens, such as medication, food, dust, mold, or animals. Information related to allergy was specifically asked to all patients and volunteers, emphasizing details related to each allergic manifestation. All reported allergies were proved by medical diagnosis, identifying allergens for each case.
The present study included 60 cancer patients, distributed in three subgroups (n = 20), gastric adenocarcinoma, prostate adenocarcinoma, and head and neck squamous cell carcinoma (HNSCC). These tumors were selected because they exhibited an inverse relation with allergies in a previous epidemiological study, within the same line of research 40 . All patients in both groups were identified by gender and age. Cancer patients were subjected to the same clinical assessment as the control group, regarding allergies and their relation to cancer, considering its histology, location, staging, and diagnostic data.
Serum cytokine dosages were obtained using the ELISA method, in accordance with manufacturer instructions. This study reviewed total IgE (ALPCO, Salem, NH, USA), IL-2 (Enzo Life Sciences Statistical analysis. Statistical analyses were conducted using the Statistical Package for the Social Sciences software, version 25 [https:// www. ibm. com/ suppo rt/ pages/ downl oading-ibm-spss-stati stics-25] 44 . The Kolmogorov-Smirnov normality test was conducted to assess the normality of each parameter. A non-parametric test was used when there was no normal data, and a parametric test was used to compare normal data averages, with their respective standard deviations. An odds ratio (OR) with a 95% confidence interval was used to evaluate the association with a risk factor, which indicates association if value 1 is not involved. The correlation among the