Comparison of an alternative and existing dbs matrix for estimation of thyroid stimulating hormone ( tsh ) in adults

Nikku Yadav. Currently, Whatman 903 (W-903) filter paper is the only filter paper that has been used for collection of blood samples in thyroid hormones. This commercially available filter paper has so expensive which is not suitable for epidemiological study conducted in developing countries. We therefore introduced a new matrix for collection of blood samples and also compared with commercial available matrix, Whatman filter paper (W-903). We used two matrices initially for standardisation purpose viz., the conventional Whatman dbs cards (#903 10531018) & pre coated TLC silica gel matrix (60F254, Merck 20x20 Cm), the latter with cost-effectiveness in view. Both were subjected to validation for optimum working conditions. The assay characteristics for Whatmann DBS card and TLC matrix were found to be as follows: correlation coefficient 0.982 & 0.976, slope 0.070 & 0.068, coefficient of variation 0.32-0.77% & 0.40-0.85%, sensitivity 0.0287 & 0.0286 mIU/L and interassay variation 0.40% & 0.50 respectively. The assay characteristics of the calibration curve for FT4: correlation coefficient 0.926 , slope 0.094, coefficient of variation 0.84-4.12% and inter assay variation 3.03% for TLC silica matrix. Drying time (0.5, 1.5, 2.5, 3.5 , 4.5 ,5.5 h), elution time (1.0, 3.0, 5.0-16 h), percent recovery (83-98%) and linearity at varying dilutions were determined. Stability studies for TSH samples were conducted at variable temperatures and half life found to be temperature dependent viz.,24.54 days, RT; 180.45, 4C; 206.25 days, -20C. Owing to its cost-effectiveness and comparable results, TLC silica matrix was preferred for subsequent studies under optimized conditions.


Introduction
Dried blood spot sampling technique has emerged as a pertinent method for both qualitative & quantitative bioanalysis context (Sharma et al., 2014). DBS has been mainly used for the neonatal screening (Pitt 2010), therapeutic drug monitoring (Edelbroek et al., 2009, Jantos 2013, preclinical (PK/TK), clinical study, metabolomic (Yin et al., 2015), disease diagnosis (Jantos 2013), detection of inborn errors of metabolism, lipidomics (Koulman et al., 2014), thyroid hormone screening in new born babies (Miyai et al., 1978, Mizuta et al., 1982, Torresani et al., 1986, Lemonier et al., 1999  This method is a sampling method for blood samples on filter paper. After initial sampling filter paper is dried, stored and taken to the laboratory for further analysis. DBS method offers distinct advantages over conventional serum sampling technique e.g use of small blood sample, ease of collection, reduced sample processing & shipping cost and various associated & ethical benefits.
Uses of commercially available DBS matrix become more costly for epidiolological surveys. In this present study a TSH levels of an adult population have been determined using novel DBS sampling method with slight modification.

Material & Method:-
Materials:-Matrix used were purchased from G E healthcare Pvt ltd (whatman 903 dbs card), Gurgaon & MM Biotech Pvt ltd (TLC Silica)New Delhi. Thyroid stimulating hormone (TSH) standards (0 to 40 mIU/ml) were purchased from cal biotech ltd, New Delhi. Disodium hydrogen phosphate, potassium di hydrogen phosphate from ficher scientific, sodium hydroxide from SRL and sodium chloride from merck were purchased. ELISA spectrophotometer & Immunowasher were purchased from BIORAD and pH meter was supplied by Chandra scientific systems Pvt Ltd. All blood samples for TSH analysis were collected from the volunteers with mixed age groups. Hormone free Serum was prepared in laboratory by activated charcoal method & washed erythrocytes were separated from blood pool.

Methods:-
Finger pricked method was used to collect blood sample (McDade 2014). Blood was applied to matrix & dried for different times. The responses were monitored by punching out 4 mm diameter spot at 0.5, 1.5, 2.5, 3.5, 4.5 & 5.5 hr intervals. These dried blood spots were eluted with different elution solvents for different incubation time. i.e millipore water, distill water & phosphate buffer saline at pH 7.4. The responses were observed using ELISA spectrophotometer. The phosphate buffer saline (Xianing et al., 2007) was prepared in millipore water & distills water. The natural pH of distill water (ph 6.8) and millipore water (pH 6.1) were adjusted to 7.4 & measured the responses. Different molar phosphate buffer saline was also prepared and measured responses.
TSH was assessed in Dry Blood Spots using a modification of Zimmerman et al., 2003 and read in ELISA plate reader (Adachi et al., 2012). Pre coated TLC silica gel matrix (60 F254 20 x 20 cm) was used for blood collection, instead of Whatman paper. In order to minimize matrix differences and maximize comparability between calibrators and unknowns, DBS standards was prepared (Mcdade 2014). Each concentration of TSH standards was then added to an equal volume of washed erythrocytes (1:2 dilutions). Mix gently with inversion for 5 min, then applied each calibrator to matrix. Blood samples were also spotted on whatman dbs card and TLC silica matrix. After drying of 3.30 hr, punched out a disc of 4 mm diameter and placed in appendorf tube. Elution solvent (110 µl) added to each tube, intermittent vortex for 1 mint and incubated for 5 hrs at 4-6 0 C. 50 µl of standard elute and sample elute added to anti-TSH antibodies coated micro wells and were incubated with peroxidase labeled anti-TSH monoclonal antibodies for 60 min at room temperature. After three time washing, the unbound antibodies were washed off and bound conjugate remains in micro well. These bound conjugates further react with substrate 3,3 ,5,5 -tetramethylbenzidine (TMB), incubated and added stop solution to each well. Read absorbance on ELISA reader at 450 nm within 15 minutes after adding the stopping solution. By comparing the OD of the test samples to this standard curve, the concentration of the TSH was determined.
Four standard curves were run on different days. Standards and controls were run in set of five. Figure depicts a standard curve obtained with TSH standard dried blood discs. The measurable range was 0.25 to 20 mIU/ml. The smallest amount detectable was 0.0287 and 0.0286 mIU/ml on whatman dbs card and TLC silica matrix from the pint where the 95% confidence limits of the response at zero doses intersects the standard curve. The intrassay variation for each run & interassay variation were calculated.
Linearity of the assay was also measured by using TSH standards with low, medium & high conc. (0.25, 2.5, 20 mIU/L) were diluted with equal amount of washed erythrocytes (300µl) to prepared artificial blood. This artificial blood with known concentration of analyte further diluted with hormone free serum (1:1.5, 1.3). Then spotted on the both the matrix (Whatman & TLC). Calculated the ratio of observed to expected analyte concentration was used as measure of accuracy.
Effect of hematocrit was assessed by preparing artificial blood with known concentration of TSH varying hematocrit (25% to 75%). Human K2EDTA whole blood was centrifuged at 4 0 C for 10 min at 1000 g. The resulting plasma & erythrocytes fractions were separated as Mac dade 2014. Hormone free serum was prepared by added 1gm norit A Charcoal to 19 ml of serum. For the preparation 30 % hematocrit, 30 µl of RBC mixed with 10 µl TSH concentrations from each calibrator and 60 µl of hormone free serum was added to make up the volume of 100 µl. Same protocol was followed for preparation of different hematocrit with changing the amount of RBC, hormone free serum and known amount of TSH remain constant at different concentration. Punched out 4 mm diameter disc & measured the TSH response through immunoassay. Recovery was also calculated.
The quality control samples were spotted on whatman filter paper and TLC silica matrix. These matrixes were air dried for 3.30 hr then processed for TSH assay in five replicates from each set. The mean value of TSH concentration was calculated. This amount of TSH was considered as zero day concentration. A set of controls were stored at three different storage condition i.e room temperature, 4 0 C and -20 0 C in air tight zip lock plastic bags with silica balls. The silica balls were regularly checked & changed if required. Controls stored under these conditions were assessed on subsequent hrs 0, 24, 96, 216, 456, 696 for TSH. Percentage of hormone degradation was calculated & half life in all sets was also determined. We used to determine one way analysis of variance to determine the any change in assay values for control stored at different condition for different hours.

Comparison of DBS and Serum:-
Blood samples were collected from subject (n=20), applied to matrix and also separated serum from similar blood sample. After drying of 3.30 hr, punched out a disc of 4 mm diameter and placed in appendorf tube. Elution solvent added to each tube, vortex and incubated for 5 hrs. DBS elute and serum added to anti-TSH antibodies coated micro wells and were incubated with peroxidase labeled anti-TSH monoclonal antibodies. After washing, the unbound antibodies were washed off and bound conjugate remains in micro well. These bound conjugates further react with substrate 3,3 ,5,5 -tetramethylbenzidine (TMB), incubated and added stop solution to each well. Read absorbance on ELISA reader at 450 nm within 15 minutes after adding the stopping solution. By comparing the OD of the DBS samples to standard curve with slope 0.068, the concentration of the TSH was determined. Similarly OD of serum sample was compare with standard curve (Slope=0.092) and TSH concentration was estimated.
The correlation coefficient for the two values of same sample (TSH in serum & DBS) was 0.95, (Pearson r= 0.97) the slope was 1.013 and the intercept was 0.256. Correlation between TSH DBS and TSH plasma concentrations supports the use of DBS for hormonal studies.

Result & Discussion:-
Dried blood spot a convenient sampling method used for collection of whole blood by heel prick method. This method is useful in resource limited area especially in the developing countries where handling samples for analysis reduced cost as well as time (Brambilla et al., 2003). The present study being describes here compare the three matrix, two alternative new DBS matrix and one commercial DBS card. The matrixes were compared for physical characteristics of DBS (spot areas) at different hematocrit and accurate quantification of analytes (recoveries). The colour of the blood spot deepens from bright red to deep red when the hematocrit level is increased from 30 to 75%. As the blood hematocrit level changes, the viscosity of the blood changes, high hematocrit levels being most viscous due to the increase in number of blood cells. The variability in viscosity leads to differences in diffusion properties of blood through different matrix, which can directly impact size of the blood spot formed. The are covered by same amount of blood on whatman filter paper was almost more than three times to the TLC silica matrix and two times to the whatman dbs card. More blood spread to the whatman filter paper and also covers more area as compared to other dbs matrixes. Figure 1 illustrate dependence of the spot diameter (mm) on the hematocrit value (%) on TLC matrix (A), whatman filter paper (B) and whatman dbs card.
The volume of blood in 4 mm punch was 50% higher on TLC matrix than whatman filter paper at all hematocrit as shown in table 1. This shows that more analytes is extracted, resulting increase sensitivity or better responses and higher recovery. The recoveries of TSH hormone on the TLC matrix are high as compared to whatman filter paper at each hematocrit level as shown in table 1. The whatman filter paper shows lesser recoveries (79%) at different hematocrit as compared to TLC silica matrix (90%) and commercially available whatman dbs 903 card (90%).
The volume taken from the 4 mm punch is determined by the ratio of the area of the area of the punch to the area of the spot. The blood volume in a 4-mm disc taken from a DBS was positively correlated with hematocrit, over the range 30 - Samples collected on TLC silica matrix shows almost twice half life as compared to whatman filter paper at various storage conditions (Table 2). Whatmann filter paper having more spreading, lesser recovery & less sample stability hence this matrix is not further recommended for the sample collection.

Assay Protocol:-
Modified DBS protocol produces valid results we need to demonstrate a reasonable level of accuracy, precision, and reliability in our laboratory measurements. It is therefore essential to identity sources of error, to minimize their impact to the extent possible, and validate each parameter for the analyte. In addition, it is important to consider the range of values that can be quantified, and the level of accuracy, precision, and reliability across this range. DBS sampling involves spotting of blood on the matrix, drying, incubation and biochemical assay.  . After drying analyte contained in a DBS sample must be brought into solution before analysis. A standard hole punch is typically used to cut out discs of whole blood of uniform size, and one is placed into an elution buffer for 5 hrs. Phosphate-buffered saline is a good place to start for an elution buffer as reported in published study (Neto and Schulte J 1998, Abdulaha et al., 1999, Hoffman et al., 2003 but present study reported less responses of TSH. In this study dried blood spots were eluted in distill water, millipore water & phosphate buffere saline (PBS). The responses observed in elution with phosphate buffer saline at pH 7.4 were significantly lower than distill water and millipore water as depicted in figure 2. Responses were increased with increase in pH of the millipore water. Responses of PBS were remains the same, when prepared in distill water as well as millipore water. Figure 3 represented that the response observed in millipore water was remains significantly higher (p<0.0001) as compared to phosphate buffer saline which was prepared in millipore water.

Figure 3:-
The response of TSH in millipore water was higher as compared to millipore phosphate buffer saline at similar pH 7.4.
The response of TSH in phosphate buffer saline cannot be enhanced whether it was prepared in distill water and millipore water. The response in millipore water remains high as compared to phosphate buffer saline after consideration of all suitable conditions.
Responses were continuously responses decreased when molarities of the PBS increased, maximum response of TSH conc. were observed at the 10 Mm PBS. In present study, millipore water was first time use for the extraction TSH from DBS matrix. Only one study (Lee et   In our study whatman 903 dbs card was used to collect the blood specimen and added with different known concentration of TSH (range from 0.25 to 20 mIU/ml). Absorbance measured at 420 nm was plotted as TSH concentration and range of TSH estimation ranging from 0.25-20 mIU/ml . The slope and regression coefficient of the standard curve was found to be 0.070 and 0.982 respectively (figure 6, 7   The present method being described here involves the use of TLC silica matrix as DBS matrix. The different conditions for the TLC matrix were optimized and standard curve presented in figure 5.

Comparison of DBS and Serum:-
The

Conclusion:-
This present assay is highly sensitive with a low blood volume requirement & low cost, it is especially attractive for special patient populations such as pregnant population who have conditions that limit the amount of allowable blood obtained for this type of study. This method does not require any buffer preparation for sample incubation. Until now, an alternative DBS matrix has not been available for laboratories to utilize throughout their various stages of processing, collection & storage and cost effective in nature.