Abstract
Patients with hereditary spherocytosis (HS) have increased rates of erythropoiesis and higher folate requirements. In a case-control study of patients with HS, we evaluated the associations between the use of 5 mg folic acid (FA) daily and serum concentrations of folate, unmetabolized folic acid (UMFA), interleukin (IL)-6, IL-8, IL-10, interferon-γ (IFN-γ) and tumor necrosis factor-α (TNF-α); and mRNA expression of dihydrofolate reductase (DHFR), methylene tetrahydrofolate reductase (MTHFR), IL8, IFNG and TNFA genes. Total serum folate and folate forms were measured in 27 patients with HS (21 users [HS-U] and 6 non-users [HS-NU] of supplemental FA) and 54 healthy controls not consuming 5 mg/day supplemental FA. Each patient was matched to two controls based on age, sex and body mass index. The mononuclear leucocyte mRNA expression of relevant genes and their products were determined. Serum folate, UMFA, 5-methyl-tetrahydrofolate (5-methyl-THF) and tetrahydrofolate (THF) concentrations were significantly higher in HS-U compared with matched healthy controls (p<0.001, n=42). HS-NU had lower serum folate concentrations than matched healthy controls (p=0.044, n=12). HS-U and HS-NU presented similar hematological and biochemical markers profiles. No differences were found between HS-U and HS-NU for cytokine serum concentrations and mRNA expression genes. DHFR mRNA expression was higher in HS-U than in HS-NU. The use of high daily doses of FA for treatment of patients with HS may be excessive and is associated with elevated serum UMFA and elevated DHFR mRNA expression. It is not known whether long-term high-dose FA use by patients with HS might have adverse health effects.
Significance of this study
What is already known about this subject?
Patients with hereditary spherocytosis (HS) present with chronic anemia and require greater amounts of folate for maintaining adequate erythropoiesis.
Most of these patients receive 5 mg/day of folic acid, a synthetic molecule that needs to be reduced to tetrahydrofolate by the enzyme dihydrofolate reductase (DHFR) before it can be be incorporated into the active cellular pool of folate.
This is a high dose of folic acid (fivefold the tolerable upper intake level of folic acid for healthy subjects), and part of this folic acid circulates in blood as unmetabolized form.
There are concerns about the risk of such high concentration, especially in relation to unmetabolized folic acid (UMFA).
What are the new findings?
In our study we demonstrate, high concentrations of UMFA and total serum folate in patients with HS.
To our knowledge, this is the first time UMFA was determined in patients with HS.
In a previous study by us, we demonstrated high folic acid could affect natural killer cell activity.
We demonstrate higher DHFR mRNA expression in patients with HS using 5 mg/day of folic acid compared with patients with HS that are not using folic acid.
We performed a hematologic and inflammatory profile in patients with HS.
How might these results change the focus of research or clinical practice?
Differently than previously thought, excess folic acid is not completely eliminated in the urine and persistently high circulating concentrations can be observed in these patients.
We believe that the doses of folic acid used routinely in patients with HS need to be reduced to minimize potential health risks.
Since determination of UMFA is not available in routine laboratories, determination of total serum folate may be useful for monitoring therapy in these patients.
Thus, we propose an upper limit showing high sensitivity, and specificity cut-off for serum folate concentration that is associated with elevated UMFA concentrations.
Introduction
Hereditary spherocytosis (HS) is an inherited hemolytic anemia most prevalent in Caucasians, affecting around 1 in 2000 individuals. The HS phenotype may arise from one of several mutations that lead to defects or deficiency in one or more proteins present in red blood cell (RBC) membranes, such as band 3, protein 4.2, ankyrin and β-spectrin.1 As a result, RBCs lose small vesicles of cytoplasmic membrane-containing lipids and proteins, reducing their membrane surface area, and leading to a spherical-shaped (spherocytes).1 The abnormal shape leads to reduced deformability of the cells and consequent premature trapping and destruction by the spleen. Shortened red cell lifespan results in increased compensatory rates of red cell production.2 3
To ensure adequate nutrient requirements for increased erythropoietic demands, folic acid (FA) is usually prescribed for patients with chronic hemolysis. However, guidelines are lacking regarding the optimal dose of FA.1 4
FA is a synthetic form of folate that must be reduced by the enzyme dihydrofolate reductase (DHFR) to be incorporated into the active cellular pool of folate.5 The activity of DHFR could be a limiting factor for FA reduction in people consuming greater than the tolerable upper intake level (UL) of 1 mg/day of the vitamin in the form of FA,6 leading to the appearance of unmetabolized FA (UMFA) in the circulation. Although no definitive adverse outcomes have been linked to high FA intakes, a recent study carried out by our group showed that intake of 5 mg of FA daily for 3 months by healthy individuals reduced the number and the cytotoxic activity of natural killer (NK) cells and was associated with elevated mRNA expression of interleukin-8 (IL8) and tumor necrosis factor-α (TNFA) genes.7
In Brazil, patients with HS and other chronic hemolytic states are being prescribed 5 mg of FA daily, which is five times higher than the UL for healthy individuals.8 It is not known what the effects of this high FA dose are in chronically supplemented individuals.
Thus, the aims of this study were to assess whether there is an association between elevated UMFA and proinflammatory cytokine levels when patients with HS were treated long-term with 5 mg of FA daily.
Methods
Twenty-seven patients with HS seen at hematology centers in the cities of Sao Paulo and Sorocaba (Brazil) were recruited for this study and 54 healthy people were recruited as a control group from the same community. Each patient was matched to two controls by age, sex and body mass index (BMI), the main covariates that may confound the analysis. Although patients were prescribed 5 mg of FA daily, six of them self-reported not taking FA during the interview conducted by our team. We took advantage of the fact that some patients were not using FA daily because of non-compliance with the treatment, which enabled us to compare patients with HS who were users (HS-U) and non-users (HS-NU).
A written informed consent was obtained from all subjects prior to their inclusion in the study. The analysis of blinded specimens by the CDC Nutritional Biomarkers Laboratory (serum folate forms) does not constitute engagement in human subject research. Other blood measurements were performed in laboratories at the University of Sao Paulo, Brazil.
The exclusion criteria were: age <18 years, chronic alcoholism, acute or chronic diseases, pregnancy, use of immunosuppressive drugs and people who donated or received blood transfusion within 6 months of the study. Furthermore, control subjects who used FA, either in the form of multivitamins or vitamin-minerals mixtures in the last 6 months were also excluded from the study.
Venous blood samples were obtained by venipuncture from each participant after an overnight fast (8–10 hours). Serum folate was determined by a microbiologic assay with the use of Lactobacillus casei (chloramphenicol-resistant strain NCIB 10463) and FA (F7876 Sigma-Aldrich FA 98%) as calibrator.9 Concentrations of serum folate forms (UMFA, tetrahydrofolate [THF], 5-methyl-THF, 5-formyl-THF, 5,10-methenyl-THF and MeFox [an oxidation product of 5-methyl-THF]) were analyzed by high performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS).10 Serum total folate was calculated as the sum of these folate forms excluding MeFox. Plasma methylmalonic acid (MMA) was assessed by HPLC-MS/MS. Other biochemical analyses, as well as hematological, mRNA expression analysis and NK cell count, were performed as described previously.7
BMI was calculated by dividing weight (kg) by height2 (m) and the measures were performed as previously described.11 Subjects were classified as underweight (BMI<18.5 kg/m2), normal weight (BMI 18.5–24.9 kg/m2), overweight (BMI 25.0–29.9 kg/m2) and obese (BMI≥30.0 kg/m2).12
Dietary intake assessment
Dietary intake was assessed by two 24 hours dietary recalls (24 hours) in each period of the study, the first one on the day of blood collection and the second one a few days after blood collection, as previously described.7 11
Statistical analyses
Statistical analyses were carried out using SPSS V.22.0 (IBM, USA) and GraphPad PrismV.5.04 (GraphPad Software, USA) software.
Serum folate concentration <7.0 nmol/L was defined as folate deficiency.13 The value of 45 nmol/L (20 ng/mL) for serum folate was considered as a cut-off due to being the upper limit of the assay calibration curve in many serum folate assays. Elevated serum UMFA was defined as a serum concentration >1.98 nmol/L, according to the 95th percentile in the distribution of values obtained for fasted persons in the National Health and Nutrition Examination Survey 2007–2008.14
Elevated plasma total homocysteine (tHcy) was defined as a plasma concentration>13.9 µmol/L.15 Vitamin B12 deficiency was defined as a serum B12concentration<256 pmol/L, while functional vitamin B12 deficiency was defined as a B12concentration<256 pmol/L and MMA>271 nmol/L.16 Anemia was defined as a hemoglobin (Hb) concentration<120 g/L for women and <130 g/L for men.17
HS disease severity was classified adapting the criteria previously described,18 by using values of Hb, reticulocyte count (Ret) and serum total bilirubin (TBi). Thus, patients with HS were stratified as trait: Hb normal, Ret<3% and TBi<17 µmol/L; mild: Hb 110–150 g/L, Ret 3%–6% and TBi 17–34 µmol/L; moderate: Hb 80–120 g/L, Ret 6%–10% and TBi 34–51 µmol/L; severe: Hb 60–80 g/L, Ret>10% and TBi>51 µmol/L. To be accepted in a category the patient needed at least two parameters compatible with the classification. When the three parameters (Hb, Ret and TBi) were not enough to establish the disease staging, Hb was used as the arbiter criterion for classification.
The Χ2 test, or the likelihood ratio and Fisher’s exact tests were used to analyze the frequencies of categorical variables. The Mann-Whitney U test was used to compare numeric variables between groups.
Spearman’s correlations between serum UMFA and serum cytokine concentrations and among serum cytokine concentrations were performed in each group studied.
In order to define a cut-off for serum folate that represents a higher likelihood of having elevated UMFA concentrations, the area under the receiver operating characteristic (ROC) curve was obtained for UMFA>1.98 nmol/L as a dependent variable and serum folate concentration measured by HPLC-tandem mass spectrometry as an independent variable. The selected cut-off maximized sensitivity and specificity.
The level of significance was set at α=5%, p≤0.05.
Results
Among the 27 patients with HS enrolled in this study, 21 were users of 5 mg FA daily (HS-U), while 6 were non-users (HS-NU). Regarding BMI, only one woman in HS-U, one woman in HS-NU and their respective controls were classified as obese. The characteristics of the patients with HS and control subjects are shown in table 1.
Natural food folate intake, FA intake and total food folate intake, expressed as dietary folate equivalents (DFEs), did not differ significantly among HS-U, HS-NU and their controls (table 2).
Serum folate, UMFA, 5-methyl-THF, THF and MeFox concentrations were significantly higher in HS-U compared with their controls (p<0.001) (table 2). On the other hand, HS-NU patients had lower serum folate concentrations (p=0.044) and did not show differences regarding UMFA, 5-methyl-THF, THF and MeFox concentrations compared with their controls (table 2). In addition, HS-U had higher serum folate, UMFA and 5-methyl-THF than HS-NU patients (figure 1).
In HS-U patients, serum UMFA represented on average 11.4% of total folate (HPLC-MS/MS), while it represented on average 2.44% in their controls, 3.97% in HS-NU patients and 2.23% in HS-NU controls. The contribution of 5-methyl-THF to serum total folate (HPLC-MS/MS) ranged from 63.3% in HS-U to >90% in HS-NU and both control groups.
No difference was found between HS-U or HS-NU patients and their controls concerning serum vitamin B12 and tHcy concentration (p>0.050). However, MMA concentration was higher in HS-U patients than in their healthy controls (p=0.003). Vitamin B12 deficiency was observed in 23.8% of HS-U patients, but the deficiency frequency was similar to their healthy controls. Even when functional deficiency of vitamin B12 was observed, there was no difference between HS-U and their healthy controls (table 2).
We found no significant difference between HS-U and HS-NU patients for any of the hematological or biochemical parameters, including the inflammation markers (table 3). Patients with HS (users or non-users of 5 mg of FA daily) presented lower RBC counts and Hb concentration and higher values of reticulocytes and serum lactate dehydrogenase activity when compared with their controls (table 3). Most HS-U and HS-NU patients were anemic (76.2% and 83.3%, respectively) compared with only 3 (5.9%) healthy control subjects. Only HS-U patients had higher white blood cell (WBC) and NK cell counts than their healthy controls (table 3). When HS-U patients were classified according to the severity of clinical presentation of the disease (trait, mild, moderate or severe), no significant differences in NK cell counts were observed.
Higher concentrations of IL-8 and TNF-α were found in HS-U patients compared with their healthy controls (p<0.001), while HS-NU patients presented higher TNF-α concentrations (p=0.028) but similar IL-8 concentrations compared with their healthy controls (table 3). No differences were observed between patients with HS and their controls regarding serum concentrations of IL-6, IL-10, interferon-γ (IFN- γ) and high-sensitivity C-reactive protein (hs-CRP) (p>0.05) (table 3).
We found no correlation between the serum cytokine concentrations with serum UMFA in 54 controls (table 4). However, there were inverse and moderate correlations (close to significance) between UMFA and IL-10 (rho=−0.400, p=0.072), and UMFA and IFN-γ (rho=−0.396, p=0.076) (table 4).
Higher DHFR mRNA expression was found in HS-U compared with HS-NU patients. No differences were found between the two groups for methylene tetrahydrofolate reductase, IFNG, TNFA and IL8 mRNA expression genes (figure 2).
The inflammatory biomarkers and mRNA expression evaluated in this study did not show significant differences when patients with HS were classified according to the severity of their clinical presentation.
The cut-off for serum folate obtained by ROC analysis was 54 nmol/L (sensitivity=100% and specificity=91.7%) and the area under the curve was 0.982 (95% CI 0.957 to 1.000, p<0.001).
Discussion
To our knowledge, this is the first study that evaluated serum folate forms and inflammatory markers in patients with HS receiving high doses of FA, exceeding the UL (1 mg/day). Our findings show that daily use of 5 mg of FA among patients with HS resulted in unusually high serum folate concentrations and its non-metabolized form UMFA compared with healthy controls not consuming supplemental FA. Furthermore, a high proportion of elevated UMFA concentrations was found in HS-U patients when compared with their controls. The data that we obtained on patients not using folate were fortuitous and a result of patient non-compliance and the inclusion of six patients with HS not using FA supplementation made it possible for us to assess the same biochemical markers when only one variable changed compared with the control group (ie, disease status), although in a small group. We observed lower serum folate concentrations in HS-NU when compared with their healthy controls, confirming that patients with HS require an additional supply of FA. However, 5 mg of FA daily may be an excessive dose and it may be more prudent to substantially reduce the daily dose. Titration of the ideal safe dose would require further studies to ensure adequacy of folate intake through monitoring of homocysteine or others metabolite status.
The appropriate FA dose is a subject of discussion for hemolytic anemias, including HS, especially according to disease severity. A few studies have described different doses, ranging from 1 mg/day, 2.5 mg/day, 5 mg once a week and even 100 µg/kg/day.4 19 20 The FA dose recommended by the Brazilian government for women in the periconceptional period is 5 mg daily,21 and this dose has also been used for patients requiring higher amounts of FA due to HS or other hemolytic anemia over their entire lives.
In a previous study, we demonstrated that the daily use of 5 mg of FA by healthy individuals over 90 days, in a non-controlled intervention, led to increased UMFA concentrations, increased mRNA expression of the DHFR gene and of the cytokine IL8 and TNFA genes, and reduced number and cytotoxicity of NK cells.7 The present study confirmed the increased mRNA expression of the DHFR gene. We hypothesize that the enzyme may be upregulated due to saturation in the presence of excess FA. However, the present study did not find higher IL8 or TNFA mRNA gene expression in HS-U compared with HS-NU patients, possibly due to limited sample size. The much higher serum IL-8 and TNF-α concentrations in the two HS groups suggest that this may be a result of the disease (HS) and not from the use of FA. To our knowledge, this is the first study that showed the inflammatory cytokine profile, including IL-6, IL-8, IL-10, IFN-γ and TNF-α, in patients with HS chronically subjected to high concentrations of FA. In the present study, NK cell counts in HS-U patients were higher than in healthy controls. This contrasts with our previous finding in which the NK counts decreased in healthy volunteers after they consumed 5 mg FA daily for 90 days.7 Unfortunately, we were not able to study sequential changes in WBC and NK number and activity in patients with HS before and following initiation of FA supplementation, nor were we able to determine NK cell cytotoxicity in patients with HS in the present study.
While we found no differences in serum vitamin B12 concentrations and functional deficiency of this vitamin between HS-U patients or HS-NU patients and their respective controls, HS-U patients had higher MMA concentrations than their controls. On the other hand, concentrations of tHcy were similar between HS groups and their controls. This observation raises the possibility that high concentrations of folate caused by intake of supplemental FA may negatively affect B12 status, as has been noted in other reports.22–24
The serum folate concentration that is associated with elevated UMFA concentrations (>1.98 nmol/L) is not yet known. We found that a serum folate cut-off of >54 nmol/L showed a high sensitivity and specificity for predicting elevated UMFA concentrations >1.98 nmol/L. This cut-off is close to 45 nmol/L, suggested previously as a high folate cut-off,13 and could be used to distinguish between individuals with normal compared with elevated UMFA, since UMFA determination is not easily available due to high cost.
The major limitation of our study is the relatively small number of patients with HS, which is due to the low prevalence of this disease in our region. However, we highlight as strengths of this study the assessment of hematological and inflammatory cytokine profiles of patients with hemolytic anemia with low-grade inflammation using chronically high amounts of FA.
From this case-control study, we conclude that daily doses of 5 mg of FA, for treatment of patients with HS was associated with elevated serum folate forms and elevated DHFR mRNA expression. We were not able to show an effect of UMFA excess on modulation of immune response in patients with HS. More studies with a larger number of patients are needed to verify possible health effects associated with the long-term use of high FA doses, which exceed the UL in patients with chronic hemolytic anemias. This may be particularly relevant in individuals with functional asplenia or postsurgical splenectomy such as might occur in sickle cell disease or HS, who are already more susceptible to infections, particularly those caused by unicellular parasitic infections.
Acknowledgments
The authors would like to thank Patricia Mendonça Amorim for measuring serum folate and vitamin B12 concentrations. The authors would also like to thank all people who participated in this study.
Footnotes
Contributors CP designed the research study, collected the samples and data, performed the laboratorial determinations, analyzed the data and wrote the paper. MRL and JFB collected the samples and data, performed the laboratorial determinations and analyzed the data. FRL performed statistical analysis and analyzed the data. BCAG and VMC performed MMA determinations and analyzed the data. GWG performed the laboratorial determinations and reviewed the paper. MSF, RC, VLBD included the patients and performed collect of clinical data. CMP and ZF performed UMFA and folate forms by LC/MS and reviewer the paper. RG designed the research study and did an expert review of the paper. EMG designed the research study, analyzed the data, performed statistical analysis and wrote the paper.
Funding This study was financially supported by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP 2012/12912-1) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq 4826412012-6 and 401586/2014-6), and in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior—Brasil (CAPES)—Finance Code 001, Brazil. JFB, MSF and EMGS have fellowships from CNPq, and MRL has a fellowship from CAPES.
Competing interests None declared.
Ethics approval The Institutional Ethics Committee approved this study (CNS 466/12, CAEE 04389512.2.000.0067).
Provenance and peer review Not commissioned; externally peer reviewed.
Patient consent for publication Not required.