Levels of bioavailable, and free forms of 25(OH)D after supplementation with vitamin D3 in primary hyperparathyroidism

The 25 hydroxyvitamin D [25(OH)D] is the major metabolite for ascertaining vitamin D status, which circulates bound to a specific carrier (vitamin D-binding protein - VDBP). A portion that circulates unbound vary according to the VDBP genotype. This study evaluates the behavior of different forms of 25(OH)D, before and after supplementation with 14,000 IU of vitamin D3, weekly for 12 weeks, in individuals with primary hyperparathyroidism and controls. Fifty-six patients with active primary hyperparathyroidism (PHPT) and 64 paired controls (CTRL), not taking vitamin D3 for the last three months, were enrolled. The genetic isotypes of VDBP were determined to calculate bioavailable and free 25(OH)D. A p < 0.05 was considered significant. There were no statistical differences in free, bioavailable, and total 25(OH)D levels between PHPT and CTRL groups at baseline. The distribution of VDBP haplotypes 1s/1s, 1f/1f, 1s/1f, 2/2, 1s/2, and 1f/2 was similar between groups. After supplementation, all three forms of 25(OH)D proportionally increased within each group, although the percentage increment was lower in the PHPT group (p < 0.05). Total 25(OH)D is better correlated with PTH in the PHPT group than bioavailable and free 25(OH)D (r = −0.41; p < 0.05). The concentrations of total, free, and bioavailable 25(OH)D were similar in both PHPT and CTRL groups, and all forms increased proportionally after supplementation, although this increment percentage was higher in the CTRL group, with a subsequent reduction of PTH and AP. Total 25(OH)D correlated better with PTH than other forms, suggesting no advantages in measuring free or bioavailable 25(OH)D in these situations.


Introduction
The metabolic pathway of vitamin D is complex, with important effects on bone and mineral homeostasis, and additional effects on cell differentiation, cell growth, and immunomodulation in several other target tissues [1] In a typical retro-regulation loop, active vitamin D 1,25-dihydroxycholecalciferol [1,25(OH) 2 D] suppresses the production of parathyroid hormone (PTH) in parathyroid cells. Normal ranges for 25(OH)D are based on the level that results in elevation of PTH values, defining secondary hyperparathyroidism. While this inverse relationship is constant in virtually all scientific studies, it is always of low to moderate magnitude, even after correcting for calcium and renal functions. Therefore, it is believed that other parameters may affect this relationship [2].
The 1, 25(OH) 2 D is considered the active metabolite of vitamin D because of its highest affinity to the vitamin D receptor (VDR), which is 1000 times higher than that of 25(OH)D [3]. However, in vitro studies showed that even 25(OH)D can modulate PTH expression [4]. Direct action of 25(OH)D on parathyroid cells was suspected when, in the parathyroid cell culture, the expression of PTH mRNA was downregulated by 25(OH)D but also with 1,25(OH) 2 D and 24,25(OH) 2 D. To obtain the PTH suppression effect, a higher concentration of 25(OH)D is required compared to 1,25(OH) 2 D [5].
Furthermore, like all steroid molecules, vitamin D circulates bound to a carrier protein-vitamin D-binding protein (VDBP)-, which, in turn, has a different affinity depending on the single nucleotide polymorphisms (SNPs) presented (rs4588 and rs7041) [6]. Approximately 0.03% of 25(OH)D is in free form, 85% is bound to VDBP, and 15% is bound to albumin. Bioavailable vitamin D consists of free vitamin D and vitamin D bound to albumin (a weak link). The free hormone hypothesis claims that only free 25(OH)D could enter cells and perform their biological activity. However, vitamin D has a peculiar behavior, as it can be internalized into the cell cytoplasm bound to VDBP by the megalin/cubilin membrane complex and perform its biological activity. This transmembrane protein complex is present in some tissues such as the kidney, placenta, and parathyroid [6].
Regarding the molecular aspects of VDBP, few studies have tested whether patients with PHPT have a different SNPs distribution of VDBP compared to controls. Genetically different VDBP SNPs could explain the 25(OH)D variation in primary hyperparathyroidism (PHPT) patients. Studies characterizing the VDBP genotype in the Brazilian population would be of utmost importance, as this may vary according to ethnic origin [7].
Primary hyperparathyroidism presents as an autonomous increase in PTH secretion and some studies have shown vitamin D deficiency in PHPT patients [8,9]. This condition tends to lead to more severe presentations and induce larger parathyroid adenomas [10][11][12][13].
Thus, this study aimed to investigate the concentrations of total, bioavailable, and free 25(OH)D in PHPT patients and controls without PHPT, before and after vitamin D 3 supplementation. The objective is to better understand the pathophysiology of the PHPT disease and verify which vitamin D measure is most suitable for this population.

Subjects
The study was conducted at the Endocrinology Unit of the Federal University of São Paulo, São Paulo, Brazil. Patients with PHPT before parathyroidectomy, and controls, all living in São Paulo, Brazil (latitude 23°S), were selected. The research protocol was previously approved by the Ethics Committee of UNIFESP (nº 0517/2018), and all individuals signed an informed consent form before being included in the study. All inclusions occurred during Winter and the last visits occurred during Spring, periods with the lowest concentrations of 25(OH)D in our population [14].
The PHPT diagnosis was based on hypercalcemia associated with high or inappropriate PTH serum levels. The paired control group included volunteers without PHPT diagnosis and not related to the individuals affected. The exclusion criteria for both groups were creatinine clearance <35 mL/min calculated by the CKD-EPI equation (Chronic Kidney Disease Epidemiology Collaboration), family history of PHPT, multiple endocrine neoplasia (MEN) syndrome, and familial hypocalciuric hypercalcemia (FHH). All participants were not under cholecalciferol replacement for at least three months before starting the study. Postmenopause status was defined as the absence of a menstrual period for more than one year.
The participants were invited for two visits to the research center. In the first visit, after signing the informed consent form, anamnesis, physical examination, and fasting peripheral blood collection were performed. Anthropometric measurements included height and weight for subsequent calculation of body mass index (BMI). Daily intake of calcium from dietary sources was estimated using a food frequency questionnaire. During the anamnesis, there were also questions about the time of sun exposure, use of sunscreen, the practice of physical exercise, use of medications that contain estrogen or other medications that may interfere with the metabolism of vitamin D. The skin phototypes were classified according to Fitzpatrick as type I: pale white skin, blue/green eyes, blond/red hair, always burns, does not tan; type II: fair skin, blue eyes, burns easily, tans poorly; type III: darker white skin, tans after initial burn; type IV: light brown skin, burns minimally, tans easily; type V: brown skin, rarely burns, tans darkly easily; type VI: dark brown or black skin, never burns, always tans darkly [15]. The presence of nephrolithiasis was based on the chart review of the clinical history, history of lithotripsy, or complementary tests that confirm the diagnosis.
At that time, participants received 14,000 international units (IU) of vitamin D 3 (two tablets of 7000IU-AdderaD3®, provided by Mantecorp-Farmasa Inc., Brazil), to be taken weekly for the following 12 weeks. During this period, individuals were contacted by telephone to ensure adherence to the medication and protocol. In the last week of treatment, they were invited for the final appointment, when fasting blood samples were collected and empty drug blisters were counted to verify adherence.
For the first phase of the study, 70 PHPT patients and 73 controls were included. In the final visit, 14 PHPT patients and nine CTRL volunteers had to be excluded due to the following reasons: in the PHPT group, five patients underwent parathyroidectomy, seven missed the final visit, and two were not compliant to the medication (<75% doses); in the CTRL group, five missed the final visit, and four were not compliant to the medication. Fifty-six PHPT patients and 64 CTRL volunteers completed the research protocol.

Laboratory tests
Venous blood and two-hour urinary samples were collected after night fasting, between 8 a.m. and 10 a.m., and refrigerated until analysis. The samples were centrifuged at a low temperature (temperature of 4°C and rotation of 2000 rpm), and serum and plasma were separated and stored at −20°C. Total calcium (tCa), ionized calcium (iCa), PTH, 25(OH)D, creatinine, phosphate (P), magnesium (Mg), and alkaline phosphatase (AP) were measured. PTH was measured with an immunometric assay (Roche, Elecsys Analyzer 2010, USA) according to the manufacturer's recommendations. The normality range was 10-65 pg/mL, intra-assay variation was 1.6%, and inter-assay variation was 2.1%. The iCa was measured with an ionspecific electrode (AVL 9180 Electrolyte Analyzer, AVL Scientific Corp., Roswell GA, USA). The VDBP was measured with a polyclonal antibody ELISA kit (R&D Systems, Minneapolis, MN, USA), with an intra-assay variation of 6.0% and inter-assay variation of 7.2%, which can identify all VDBP subtypes. We preferred polyclonal assays once their results were confirmed by direct measurement by an LC-MS/MS (Liquid Chromatography with tandem mass spectrometry) method. Thus, the different VDBP alleles with different affinities for vitamin D metabolites do not affect the results of the immunoassay; this confusion can occur when a monoclonal antibody assay is used [16,17] The tCa, albumin, and creatinine were measured in a Cobas™ platform. The 25(OH)D was measured using a chemiluminescence immunoassay (Roche, Elecsys 2010 Analyzer, USA), with an intra-assay variation of 8.5% and inter-assay variation of 9.2%.

Total DNA extraction and VDBP genotyping
The VDBP genotype was analyzed with DNA extracted from peripheral leukocytes of PHPT patients and controls, and commercial kits were used according to the manufacturer's recommendations (Gentra Puregene Blood Kit, PUREGENE, QIAGEN) [18]. Total DNA was quantified by measuring absorbance at 260 nm in a spectrophotometer (NanoVue Plus GE Healthcare, Buckinghamshire, UK).
The free 25(OH)D can be calculated by the formula postulated by Bikle et al. [20]: where Total 25(OH)D is the concentration in ng/mL, Kalb is the constant of affinity of albumin, Kvdbp is the constant of affinity of VDBP according to the genotype, Alb is the albumin concentration in mg/dL and VDBP is the serum concentration in μg/ml. The obtained concentration of free 25(OH)D is in pg/mL. The concentration of the bioavailable 25(OH)D was obtained through the following formula:

Statistical analysis
This study used the Shapiro-Wilk normality test. The quantitative data are presented as mean ± standard deviation (SD), median, or percentage. Non-parametric tests were used to analyze non-Gaussian variables. One-way analysis of variance (ANOVA) was used to compare the means of the groups. Pearson's chi-square test (χ2) was used to test categorical variables. All analyses were performed in the GraphPad Prism Statistics, version 8.0.1 (Chicago, IL, USA). The values of p < 0.05 were considered significant.

Demographic and clinical characteristics
The demographic and clinical characteristics of the 56 PHPT patients and 64 controls are presented in Table 1.
The groups were similar for sex, age, menopause, and skin color distribution (Fitzpatrick). As expected, the PHPT group had a higher number of individuals with nephrolithiasis, previous or current use of bisphosphonate, and osteoporosis. The biochemical data showed higher levels of PTH, tCa, iCa, CaU/CrU, AP and lower P in PHPT patients. The concentrations of total, bioavailable, and free 25(OH) D; VDBP; magnesium; and CrCl were not different between the groups at baseline (Table 1). Among the menopausal individuals, in the PHPT group 7 women underwent hormone replacement therapy (estrogen or tibolone), on average for 2 years, but all had already stopped for at least 1 year; in the CTRL group, 4 women underwent hormone replacement therapy, also on average for 2 years, none of them were doing it at the time the research started.
Frequent use of sunscreen was described by 18 (32.1%) individuals in the PHPT group and by 17 (26.6%) individuals in the CTRL group. Research participants rarely exposed themselves to the sun, when they did it was only in the morning, and the exposed areas were the face and arms. The distribution of the combined rs4588 and rs7041 (Gc1f-1f, Gc 1f-1s, Gc 1s-1s, Gc 1s-1s, Gc 1f-2, Gc 1s-2 and Gc 2-2) polymorphisms was also not different between the groups (Fig. 1).

Biochemical findings
The supplementation with 14,000 IU/week of vitamin D 3 did not cause any significant changes in total, ionic calcium, or CaU/CrU in both groups. A significant decrease in PTH and AP levels and an increase in VDBP levels were seen only in the CTRL group ( Table 2).
At baseline, the PHPT group showed significant negative correlations between PTH and all the circulating forms of 25(OH)D: total (r = −0.41, p = 0.0017), bioavailable (r = −0.32, p = 0.02), and free (r = −0.33, p = 0.01) ( Fig. 3), which disappeared after supplementation. The CTRL group did not show a correlation between PTH levels and total, bioavailable, and free 25(OH)D at baseline and after supplementation. However, in this group, there was a significant decrease in PTH levels after supplementation, not related to total, bioavailable, and free 25(OH)D before or after vitamin D 3 supplementation, nor with the increment of any 25(OH)D form.
Overall, the oral use of vitamin D 3 was well tolerated in both groups, without adverse events.

Discussion
Primary hyperparathyroidism (PHPT) patients were clinically and biochemically different from the control group, as expected. They have a higher total and ionic calcium and Values were considered significant if *p < 0.05 Fig. 2 Increment of the total, bioavailable, and free 25(OH)D in the PHPT e CTRL groups after cholecalciferol supplementation. Analysis of the elevation of the total, bioavailable, and free vitamin D in each group (PHPT and CTRL) after supplementation with 14,000 IU of vitamin D 3 for 12 weeks. There was an increase in 25 vitamin D forms in both groups but there was a higher increase in the CTRL group than the PHPT group for total 25(OH)D PTH, with lower phosphorus and a higher rate of osteoporosis and nephrolithiasis. However, the total, bioavailable, and free 25(OH)D levels were similar in both groups at baseline (p = 0.10, p = 0.11, and p = 0.17, respectively). In this current study, only individuals without vitamin D supplementations in the last 3 months at least were included. Factors that may interfere with the 25(OH)D value such as age, gender, skin pigmentation and BMI are similar between groups. In addition, all participants collected blood samples during winter and resided in the city of São Paulo, that is, with the same latitude. Some researchers found lower concentrations of total 25(OH)D in PHPT patients than in controls [21], whereas others did not [22,23], suggesting the possibility of regional differences. Despite population differences, only a few articles are concerned with vitamin D 3 supplementation before starting the studies or with the season of blood collection.
To measure VDBP concentrations we chose a polyclonal assay, which recognizes all the predominant VDBP molecules defined by their respective genotypes. This is an important detail, especially in a very miscegenated population like ours. Some assays using a monoclonal antibody may not recognize very well all different VDBP phenotypes, leading to misinterpretation of the results, as described elsewhere [16,17] VDBP isotypes are characterized by polymorphisms rs7041 and rs4588, which generate different electrophoretic mobilities [24]. These polymorphisms can also trace ancestry, mainly African versus European, through amino acid determinations [24]. The ethnic origin can determine the occurrence of VDBP polymorphisms, and African, African-American, and Asian populations with darker pigmentation are more likely to carry the VDBP 1f variant, while the VDBP 1s form is more frequently observed in white populations [25]. The Brazilian population has a wide range of ethnic backgrounds, with a high degree of miscegenation, which may vary regionally. The population of the city of São Paulo, Brazil, where our study was conducted, is highly miscegenated, especially between European, African, and Asian descendants. This could be seen in the distribution of VDPB, with a higher proportion of the 1f variant and was similar to another analyzed population from southern of Brazil [26]. Other authors did not show differences in the distribution of the VDBP genotypes (rs7041 and rs4588) in more homogeneous populations [25]. The data in the present study did not show differences in the distribution of VDBP genotypes between PHPT and CTRL groups, in agreement with other authors [21,27].
Vitamin D 3 supplementation is recommended to prevent postoperative hungry bone syndrome in PHPT patients [28]. In the present study, a dose of 14,000 IU of vitamin D 3 weekly, for 12 weeks, was safe and without significant changes in both serum and urinary calcium, despite the 40.7% increase in total 25(OH)D concentrations. A metaanalysis with 10 interventional studies evaluating the effects of vitamin D supplementations compared with placebo in 340 PHPT patients demonstrated a significant increase in 25(OH)D levels, followed by a reduction in serum PTH levels, without changes in serum calcium and urinary calcium compared to the baseline [29]. In the present study, a significant decrease in PTH was seen only in the CTRL group. A longer supplementation period may be required to reduce the PTH in PHPT patients.
Some authors described lower concentrations of VDBP in PHPT patients than in controls, even at baseline, which could not be demonstrated by us. However, similar to our findings, when these groups were supplemented with vitamin D 3 , VDBP increased only in the control population, while it usually did not change in PHPT patients [21,27,30]. Bikle et al. [31] also observed that variations in VDBP genotype affect the concentration of total 25(OH)D but not bioavailable or free 25(OH)D. The present study did not show differences in baseline VDBP levels between the groups. However, after supplementation there was a significant increase of 5.6% in the CTRL group, not seen in the PHPT group. At the end, the mean of VDBP concentrations was 14.6% higher in the CTRL group compared to PHPT. There are factors already known to increase VDBP, such as estrogen replacement and pregnancy [32], and factors that decrease VDBP such as urinary protein loss [33] and malnutrition [20]. None of these factors were seen in any of our individuals. Previous studies [21,34] also found reduced VDBP in PHPT compared to controls, but the etiology remains still unclear. Some authors suggest that high levels of PTH inhibit VDBP production, as they found a negative correlation between VDBP and PTH [34]. This theory is also confirmed by Wang in a more recent study [35], which showed that after parathyroidectomy, individuals with Fig. 3 Correlations between PTH and total, bioavailable, and free 25(OH)D concentrations in the PHPT group, at baseline. There is an inverse correlation of PTH to 25(OH)D forms in the PHPT group PHPT had an increase in VDBP with normalization of serum calcium and PTH.
The present study showed that total 25(OH)D increased proportionally more in the CTRL group (61.8%) than in the PHPT group (40.7%), but the increment was not different between both groups for free and bioavailable 25(OH)D. Using a higher daily dose for a longer period (2800 IU daily for 25 weeks), Shah and colleagues reached an even higher increase (88%) in 25(OH)D levels [36]. In that study, supplemented PHPT patients showed a 17% drop in PTH [36]. Although our protocol showed a significant decrease in PTH only in the CTRL group, there was a decrease in alkaline phosphatase (AP) in both groups, which could mean a reduction in the effects of PTH in the bone tissue.
Total 25(OH)D is used in clinical practice to assess vitamin D status without distinction between 25(OH)D circulant forms. A current theory based on other endocrine systems questioned whether only the free hormone could exert biological activity. However, regarding vitamin D, the megalin/cubilin transmembrane structures in tubular kidney cells allowed internalizing the 25(OH)D-VDBP complex for further metabolization. The remaining question relates to the effects of vitamin D molecules in other tissues, beyond renal cells. Megalin and cubilin were also identified in parathyroid cells and, according to the rationale of our findings, it seems that total 25(OH)D is more relevant than the other forms for controlling PTH secretion. According to our results, the relationship between vitamin D forms and PTH in the PHPT group was most significant with total 25(OH)D than the other forms. Therefore, the present study did not show benefits in measuring either free or bioavailable 25(OH)D over total 25(OH)D in individuals with or without PHPT, corroborating other authors [36].
The limitations of the present study include the lack of information on sun exposure, although all procedures were performed during the winter/spring when the lowest levels of 25(OH)D were found in our population [14]. Moreover, there was a relatively small number of participants in each group. However, it is worth noting that the protocol was performed in PHPT patients with the overt disease, mostly symptomatic, while they were waiting for surgery, which is not easily grouped. Another strength of the present study was the absence of a degree of kinship between PHPT and CTRL groups, meaning that genetic background (VDBP genotype) was not shared. Additionally, vitamin D 3 supplementation was performed with controlled adherence, excluding those with poor compliance from both groups.
In conclusion, the distribution of VDBP genotypes was not different between PHPT and CTRL groups. This study also showed that, in individuals with overt PHPT, the supplementation with 14,000 IU of vitamin D 3 weekly for three months was safe and proportionally increased all the three circulating forms of 25(OH)D (total, bioavailable, and free), although it did not change PTH or VDBP levels. In individuals without PHPT, the increment of total 25(OH)D concentrations to values higher than 30 ng/mL significantly decreased PTH and alkaline phosphatase levels, while increasing VDBP concentrations. Finally, our results did not obtain any additional information in dosing either free or bioavailable 25(OH)D compared to total 25(OH)D in these populations.

Compliance with ethical standards
Conflict of interest The authors declare no competing interest. Our article analyzes the usefulness of free 25(OH)D in individuals with primary hyperparathyroidism before and after supplementation with cholecalciferol. There are few studies in the literature on this topic, in addition we can contribute to clinical practice by verifying whether free 25(OH)D is necessary to evaluate the patient with primary hyperparathyroidism.