Low Levels of High-density Lipoprotein in Patients with Pediatric Cancer at Diagnosis

Introduction: Cancer patients have metabolic imbalances due to the disease, treatment and their complications. Increases in triglycerides and glucose profile and catabolism of protein have been described and are associated with inflammatory response as result of the tumor activity or necrosis. Objective: To evaluate the blood lipid, lipoproteins, glucose and albumin levels in pediatric cancer patients at diagnosis. Method: Observational cross-sectional study. The inclusion criteria were children and adolescents with newly diagnosed malignancies and blood analysis results. The exclusion criteria were: previous anticancer therapy or surgical treatment, blood sample not collected and patients who refused to participate in the study. Results: It were evaluated 81 children and adolescents with newly diagnosed malignancies. There was decrease of 56% and 41% of high-density lipoprotein cholesterol (HDL-C) and triglycerides (TG), 14% and 10% of albumin and glucose, and 10% and 7.6% of total cholesterol (TC) and low-density lipoprotein cholesterol (LDL-C), respectively, for the patients. HDL-C showed statistical differences between solid and hematological cancer patients (p < 0.05). The means and medians of albumin, glucose, HDL-C and TG demonstrated that hematological patients are more prone to metabolic disturbances. HDL-C in this group was 24 ± 12 versus 40 ± 15mg/dl in other cancers. Conclusion: The primary result found in the present study was in HDL-C, mainly in patients with hematological and metastatic cancer. These disturbances could be associated with cancer-related acute inflammatory response.


INTRODUCTION
Cancer patients may present metabolic unbalances because of the disease, the treatment and their complications. Increases in lipid, lipoprotein and glucose profile and catabolism of protein have been described in these patients and are associated with inflammatory response due to the presence of the tumor, infectious or anticancer therapy complications. This condition has been described as anorexia-cachexia syndrome 1,2 , which is generated by elevated pro-inflammatory cytokines levels [3][4][5] . It is believed that inflammatory response is an important factor in the inter individual variability of response and toxic effects of cancer chemotherapy 6 .
Low HDL-C and high glucose and triglycerides levels were reported in children with leukemia and lymphoma and other forms of widespread malignant disease [7][8][9][10] . High triglycerides and glucose levels, for example, are deleterious to the immune system, increasing the risk of infection; low clearance of triglycerides leads to a partial blockage of both macrophage and granulocyte function, increasing the risk of infectious complications and morbidity [11][12][13][14][15][16] . Lipids and glucose intolerance impair oxidation and incorporation of the nutrients during inflammatory response [14][15][16] .
All these alterations have been documented in the international literature, but have never been reported in Brazilian pediatric cancer patients. Therefore, the impact of these factors in Brazilian patients is still unknown, since genetic and dietary profiles may play a role in the aforementioned alterations.
The importance of this study is to identify those abnormalities in different groups of Brazilian patients with cancer, in order to propose studies about prognosis and immunomodulation in the future. To that end, the aim of the study was to evaluate the blood lipid, lipoproteins, glucose and albumin levels in pediatric cancer patients at diagnosis.

METHOD
This is an observational cross-sectional study conducted at the Pediatric Oncology Institute (IOP) of the Federal University of São Paulo (UNIFESP-EPM) with patients who presented newly diagnosed malignancies. All the patients with previous anticancer treatment or history of corticosteroid use were excluded from this analysis.
Blood samples were collected after an overnight fast (8 hours) at the enrollment day, in ethylenediaminetetraacetic acid (EDTA) or heparin anticoagulating tubes to be immediately centrifuged. Total plasma cholesterol (CH) was measured by calorimetric method (HDL cholesterol enzymatic, K015, Bioclin, Belgium). Total Cholesterol Assay Kit measures the total cholesterol within serum, plasma, lysate, or tissue samples. The assay is based on the enzyme driven reaction that quantifies both cholesterol esters and free cholesterol. Cholesterol esters are hydrolyzed via cholesterol esterase into cholesterol, which is then oxidized by cholesterol oxidase into the ketone cholest-4-en-3-one plus hydrogen peroxide. Then, the hydrogen peroxide is detected with a highly specific colorimetric probe. Horseradish peroxidase catalyzes the reaction between the probe and hydrogen peroxide, which bind in a 1:1 ratio. Samples are compared to a known concentration of cholesterol standard in a 96well microtiter plate format. Samples and standards are incubated for 45 minutes and then read with a standard 96-well colorimetric plate reader.
HDL-C determination was performed by calorimetricenzymatic method (HDL cholesterol enzymatic, K015, Bioclin, Belgium). It is measured in serum or plasma after selective precipitation of LDL and VLDL. In a centrifuge tube pipette 250 μL of the sample and 250 μL of the precipitate, shaken strongly for 30 seconds, centrifuged at 3500 rpm for 15 minutes to obtain a clear supernatant, pipette the clear supernatant immediately after centrifugation, taking care to not resuspend the precipitate to avoid falsely high results. Colorimetry: identify the test tubes and pipette, homogenize and incubate in a water bath at 37° C for 10 minutes. The water bath level should be above the level of the reagents in the tubes, make the photometric readings of the Standard (AP) and the Test (AT) at 500 nm, resetting the device to White. The color is stable for 60 minutes.
Plasma triglycerides (TG) were enzymatically measured. Serum was separated up to 2 hours after collection and stored under refrigeration at 2 to 8° C, to maintain stability for 72 hours. The temperature of the water bath or thermostat was adjusted to 37° C. The temperature remained constant during the test. Homogenization and immediately incubation for 5 minutes at 37° C. The absorbance was read within 60 minutes against the white reagent. For calculations: triglycerides (mg/dL) = Sample Absorbance X Standard Concentration / Standard Absorbance. Calibration Factor (Fc) = Standard Concentration (mg/ dL) / Standard Absorbance. Triglycerides (mg/dL) = Sample Absorbance x Fc).
The Friedewald et al. 17 equation was applied to calculate low-density lipoprotein cholesterol (LDL-C) from CH, TG, and HDL-C Glucose Hexokinase II determined blood glucose. It is an enzymatic method that uses the enzymes hexokinase and glucose-6-phosphate dehydrogenase. Glucose is phosphorylated by ATP in the presence of hexokinase. The glucose-6-P is oxidized in the presence of glucose-6 phosphate dehydrogenase, resulting in the reduction of NAD to NADH, whose absorbance is measured. GLUH applied to biochemical systems of Advia is a reagent containing two components, reagent 1 (R1) and reagent 2 (R2). The blood sample is added to the R1, which contains buffer, ATP and NAD. The absorbance readings of the sample in R1 are obtained and used to correct the interfering substances in the sample. Then, R2 is added, initiating the conversion of glucose and the development of absorbance at 340/410nm. The difference between the absorbance at R1 and R2 is proportional to the glucose concentration.
Turbidimetrical method was used to measure serum albumin. This method is based on the dispersion of suspended particles. When a collimated beam of light reaches a suspended particle, portions of the light are absorbed, reflected or dispersed by the solution. The amount of light transmitted in the front direction is detected. The amount of light absorbed by a suspension of particles depends on the concentration and size of the particles. Solutions that require quantification by turbidimetric methods are carried out by means of photometers or spectrophotometers of visible region.
Reference of lipids was based on the "Guidelines of the National Cholesterol Education Program" 18 . When albumin and glucose were below 3.5 mg/dl and above 110 mg/dl, respectively, they were considered abnormal. It were also assessed the family history of hypertriglyceridemia and hypercholesterolemia.
Patients over 18 years old and the parents or guardians of patients below that aged signed an informed consent approved by the Institutional Review Board of UNIFESP, which reviewed and approved the study (number 0840/06).

STATISTICAL ANALYSIS
Descriptive data are expressed as means and standard deviation (SD) or median (minimum e maximum) in case of skewed distributions, and as frequencies and percentages or ranges.
The Chi-square test was adopted to compare the rate of abnormalities between hematological versus solid and metastatic versus non-metastatic cancers. T test or Mann' Whitney were used to compare means or medians for hematological versus solid and metastatic versus nonmetastatic cancer, depending on the normality of the sample. A p-value below 5% was statistically significant.

RESULTS
Eighty-one patients were included: twenty-one with hematological and sixty-four with solid cancers. The median age was 8.7 years (range, 0.2 -21.9 years). Serum TG, CH and HDL-C were evaluated in 81 patients, whereas serum albumin, glucose and LDL-C in 73, 77 and 79, respectively. Table 1 shows the demographic characteristics. Seventeen (21%) of the patients had a family history of hypertriglyceridemia and/or hypercholesterolemia; two of them had solid tumor.
The abnormalities of HDL-C rates were statistically higher (p = 0.001) in the hematologic tumor when compared to the solid tumor patients. The percentage of patients with albumin and TG abnormalities were higher in the hematologic tumor group, although not reaching statistical significance (Figure 1).
The mean of HDL-C levels of patients with hematological tumors demonstrated to be lower than the solid tumors ( Table 2).
The analysis for metastatic (n = 27) versus nonmetastatic (n = 54) cancer ( Figure 2) demonstrated higher incidence of abnormalities in the HDL-C and lower mean of the values (Table 3).

DISCUSSION
This study confirmed in Brazilian patients similar abnormalities observed in other studies 5,7-9 , mainly higher prevalence of HDL-C in hematological and metastatic cancer patients. Other studies also reported low HDL-C in patients with leukemia and lymphoma, corroborating our results. In one of these studies, abnormalities with other forms of widespread malignant disease were observed in children at diagnosis, but not in localized disease 7 . In our analysis with metastatic patients it was also demonstrated higher percentage of HDL abnormalities. Prevalence of metabolic alterations is, therefore, probably associated to the severity of the disease, when the body prioritizes protein synthesis due to the pro-inflammatory response caused by the most aggressive cancer disease. However, other mechanisms must be explored. On the other hand, glucose and albumin did not demonstrate an important decrease in the present study.
Iqbal et al. 5 showed lower mean CH, LDL-C, HDL-C and albumin in cancer in comparison to non-cancer patients; they were also significantly lower in those with metastatic solid tumors. They also found that TG was lower in non-metastatic disease and no association was observed between LDL-C, HDL-C and TG and the stage of disease 5 . In the current study, TG abnormalities were observed mainly in metastatic group, however without statistical differences in the prevalence rate. Possibly, the qualitative analysis (percentage of abnormalities) and the   Values are expressed as mean ± standard deviation, excepted for the values for triglycerides, which are also expressed as the range and the median. HDL-C = highdensity lipoprotein cholesterol; LDL-C = low-density lipoprotein cholesterol. *T test or Mann' Whitney test.  heterogeneity of the diagnosis of the sample did not allow differences to be found. Lower mean of HDL-C was found, comparing our patients with a study performed in healthy Brazilian children and adolescents [19][20][21] . Higher TG and lower LDL-C and CH, also were encountered in children with hematological cancer and metastatic patients when compared to the Brazilian study 21 .

Serum/plasma
Malignant tumors are associated with synthesis of pro-inflammatory mediators -TNF-a, Interleukins 1 and 6 and others 22,23 . These cytokines stimulate acute inflammatory response, increasing positive and decreasing negative acute phase proteins synthesis. They cause metabolic disturbances such as higher fat utilization, glucose and lipid intolerance, leading to a catabolic status 2,24 . HDL-C could be reduced due to cytokines action, acting as a negative acute phase protein.
These abnormalities are present in several conditions such as inflammation, infection and trauma [25][26][27][28] , demonstrated that elevated baseline C reactive protein levels are associated with shorter survival and development of second cancers in patients with chronic lymphoid leukemia 29 .
Albumin has been studied and high TG levels were observed in patients with blood or bone marrow invasion, suggesting a correlation with the turnover of the malignant proliferation 30 .
Malignancy also seems to influence CH levels. Hypocholesterolemia has been observed in various forms of cancer at diagnosis, particularly in hematological ones 5,31 . Low CH has been associated with increased mortality from cancer in several observational and clinical studies 32 .
Naik et al. 33 found that low CH and HDL-C together with hypertriglyceridemia were the most prominent features in pediatric patients with leukemia and Hodgkin's disease 33 .
Recently, two other studies about leukemia showed that these abnormalities are correlated with the disease activity, being reversed during remission 34,35 .
It were not observed any important abnormalities in CH and LDL-C, but apparently, the means values were slightly reduced in hematological and metastatic cancer.
Previous studies have demonstrated that cancer patients often present aberrant blood lipid profiles. Moreover, blood lipid levels of cancer patients have shown to be associated with pathogenesis and progression of cancer. High-density lipoprotein cholesterol has antiinflammatory properties.
This study has some limitations, for example, its crosssectional design, which does not allow causal analysis, the greater number of solid than hematological cancer, the absence of control group with healthy population, and the loss of some blood analyses. Future studies need to investigate subtypes of lipoprotein, tracking changes in values and analyze how they correlate with clinical outcomes. It was not possible to affirm that the present results found were related to inflammatory response because it were not measured CRP or other inflammatory marker.
Limited published studies have focused on pediatric cancer inflammation and its metabolic consequences. However, inflammation could be associated with a pre-cachexia status in a high percentage of patients at diagnosis, mainly in those with widespread metastatic disease. This condition leads to anorexia, weight loss, insulin resistance and increased destruction of muscle proteins and reduced functional capacity and quality of life [36][37][38] , increasing morbidity and mortality in cancer patients 39 . The presence of malignant tumor cells stimulates the host to produce cytokines and other pro-inflammatory mediators responsible for alterations in substrate metabolism. In some studies with cancer patients, albumin has been suggested as a prognostic factor and its alterations may be associated with other aspects, in addition to the inflammatory response 40,41 . This profile depends on the cancer diagnosis and stage/ severity of the disease 36 .
A recent study on the HDLs of young ALL survivors indicates an altered metabolism and a shift in their proteome affecting specifically their anti-thrombotic, antiapoptotic and anti-inflammatory capacities. These HDLs have an abnormal composition with a lower content of free and esterified cholesterol compared to controls. They also expressed some proteins not detectable in the control samples; most of them are either pro-thrombotic or proapoptotic. These findings may be relevant to the field of cancer survivorship in providing potential diagnostic, prognostic or therapeutic biomarkers of HDL-C functionality and metabolism in ALL 42 .
More studies are necessary to know the metabolic behavior of these components, comparing with healthy subjects, evaluating patients prospectively in different phases of treatment, and correlating with the disease, drugs, energy expenditure and nutritional condition. These abnormalities could be involved in the response to chemotherapy and in the mechanism of oncogenesis. Thus, it is important to elucidate that influence on the cancer prognostic 29,34,35 .

CONCLUSION
The primary result found in the present study was low HDL-C levels at diagnosis of cancer. Albumin and TG also presented abnormalities in cancer patients. Hematological and metastatic cancer demonstrated to influence these abnormalities more importantly. The prevalence of metabolic alterations is, therefore, probably associated to the spread and severity of the cancer, when the body prioritizes protein synthesis due to pro-inflammatory response caused by the most aggressive tumor.

CONTRIBUTIONS
Adriana Garófolo participated of the conception and design of the study, collection, analysis and interpretation of the study data, wording and critical review with intellectual contribution and approval of the final version for publication. Priscila dos Santos Maia-Lemos participated of the conception and design of the study, collection, analysis and interpretation of the study data and approval of the final version for publication.

DECLARATION OF CONFLICT OF INTERESTS
There are no conflict of interests to declare.

FUNDING SOURCES
None.