Cascade Screening of β-Thalassemia in an Indian Family Using Flow Injection Analysis–Triple Quadrupole Mass Spectrometry: Comparison of Micro Sampling Approaches with Conventional Electrophoresis

Abstract

Background: β-thalassemia is a rare genetic disorder affecting 1–5% of the global population and poses a health burden due to migration of individuals from endemic regions. Identifying asymptomatic β-thalassemia carriers is essential to prevent the birth of thalassemic babies. A simple, sensitive method compatible with self-sampling could enhance the detection of β-thalassemia in the population. Methods: Capillary blood was collected via dried blood spot (DBS) and dried blood matrix (DBM) from 18 members (52.9%, 18/34) of a three-generation family. Hemoglobin was extracted, and globin chains were analyzed on a triple quadrupole mass spectrometer (TQMS). δ/β (%) was utilized as a biomarker to identify β-thalassemia. Venous blood collected from positive and negative individuals (n = 11) was further tested to confirm the findings and validated with complete blood count (CBC) and Capillary Electrophoresis (CE). Results: β-thalassemia was detected in seven individuals: three from generation I, three from generation II, and one from generation III. CBC showed thalassemia indices, while CE demonstrated elevated HbA2 consistent with β-thalassemia. Molecular sequencing of two samples confirmed the heterozygous c.92 + 5 G > C mutation in the β-globin gene. The overall prevalence of β-thalassemia in the family was 20.6% (7/34). High clinical performance was achieved across sample types, with 100% sensitivity for DBS, 100% specificity for DBM, and an overall accuracy of 91% when compared with CE. Conclusions: TQMS in combination with CBC parameters successfully identified asymptomatic heterozygous β-thalassemia carriers using self-sampling techniques. Cascade screening within affected families emerges as a possible strategy for early detection of β-thalassemia pending comprehensive validation.

1. Introduction

β-thalassemia (β-thal) is a genetic disorder caused by the absence (β⁰) or reduced (β⁺) expression of β globin genes. Depending on clinical severity, β-thal is classified into β-thal major, β-thal intermedia and β-thal minor. Patients with β-thal major require life-long blood transfusions due to the early onset of symptoms, while individuals with β-thal minor are asymptomatic and do not require genetic counseling. The clinical symptoms of β-thal intermedia are dynamic, requiring occasional intermittent transfusions if they develop complications with β-thal [1]. The survival of severely affected individuals into adulthood can be achieved through appropriate blood transfusion and chelation practices, along with early detection of complications [2].

The global health burden of hemoglobinopathies continues to rise each year [3]. Data from the 1000 genome project indicate that approximately 8.3% of healthy individuals worldwide carry a heterozygous mutation in the HBB gene [4]. In India, the prevalence of β-thal varies widely, ranging from 0 to 10.5% across different cities [5]. Several large-scale screening programs have been initiated to reduce the incidence of thalassemic births [6,7]. Successful implementation of carrier screening programs in Mediterranean countries, such as Cyprus, Greece and Italy, has effectively prevented the birth of affected newborns [8,9].

The conventional sampling technique for the diagnosis of hemoglobinopathies is venous whole blood (WB) collection, which requires a relatively high sample volume (0.1–1.0 mL), a trained phlebotomist, and a visit to the laboratory, thereby increasing costs and logistical challenges [10]. To overcome these limitations, alternate micro sampling techniques, such as dried blood spot (DBS), are being actively explored for the analysis of various clinically relevant analytes like glucose, cholesterol, iron, potassium, albumin, vitamin B12, urea, and therapeutic drug levels, as well as in large-scale epidemiological studies [11,12,13]. DBS sampling requires a smaller volume of blood (<100 µL), minimal sample processing, and offers several advantages over conventional venous blood collection [14]. Introduced by Guthrie and Susi in 1963 for the detection of phenylketonuria in newborns [15], DBS methods have since been extended for screening sickle cell disorders [16]. The hematocrit effect on the analyte concentration is case-dependent, with no significant influence observed in the measurement of Vitamin D and Cyclosporine A [17,18]. However, Möller et al. [19] reported that elevated hematocrit levels (45–60%) may affect sample extraction efficiency.

Volumetric absorptive micro sampling (VAMS), a quantitative bioanalytical technique, enables accurate and precise collection of small blood volumes (~10 µL) using a porous hydrophilic tip [20]. Across a hematocrit range of 20 to 70%, VAMS demonstrated less than 5% variation in absorbed volume, compared to approximately 30% variation observed with DBS [21]. While VAMS offers clear advantages for quantitative analysis, both DBS and VAMS must be completely dried to prevent bacterial growth and ensure analyte stability and extraction efficiency during analysis [22].

Another emerging micro sampling device, the Biosampler, facilitates rapid drying and is used for quantitative bioanalysis of drugs [23]. The processes of collection, processing, storing and transportation associated with micro sampling devices have been comprehensively reviewed by Baillargeon and Mace [24]. Such devices provide opportunities for self-sampling in remote or resource-limited settings.

To fully exploit micro sampling techniques, the use of highly sensitive analytical instruments capable of detecting analytes from low sample volumes is essential. This integration enhances the overall analytical and diagnostic potential of these tools. Newborn screening programs for inherited metabolic diseases routinely employ mass spectrometry in combination with DBS sampling [16,25,26,27]. DBS–mass spectrometry (DBS-MS), known for its high sensitivity, specificity and analytical speed, was initially developed for small-molecule analysis in pharmacokinetic and toxicokinetic studies [28]. Its application has since expanded to the analysis of peptides and proteins, focused on hemoglobin (Hb) and its variants, for identifying clinically significant mutations [29,30,31,32]. Furthermore, rare and compound Hb variants are often detected in adult WB during HbA1c screening using triple quadrupole MS (TQMS), a low-resolution MS platform commonly available in clinical laboratories [33,34].

For a definitive diagnosis of β-thal in adults, a comprehensive approach involving complete blood count (CBC), peripheral smear evaluation, capillary electrophoresis (CE) or cation exchange high-performance liquid chromatography (CX-HPLC), and clinical correlation is essential. While CE and CX-HPLC analyze Hb as a complex, MS measures the masses of individual globin chains prior to Hb assembly. The ratio of synthesized globin chains serves as a biomarker for the assessment of β-thal [35,36,37].

The present study aims to apply micro sampling techniques to identify β-thal carriers in a family with an affected individual through cascade screening using TQMS. In addition to carrier identification, the diagnostic and clinical performance of MS-based sampling approaches (DBS, dried blood matrix (DBM) and venous WB) for β-thal detection were evaluated by comparison with established laboratory methods, that includes CBC and CE.

2. Materials and Methods

2.1. Materials and Reagents

Venous blood was collected using K₂ Ethylenediaminetetraacetic acid (EDTA) vacutainer tubes (BD Biosciences, Milpitas, CA, USA). DBS cards (Whatman™ 903) were sourced from Kala Scientific Inc. (Chennai, India). DBM devices (iOTA BioSampler™, 30 µL, iOTA Diagnostics Pvt Ltd., Gandhinagar, India) were kindly provided by iOTA Diagnostics Pvt. Ltd. (Gandhinagar, India). HPLC-grade acetonitrile and 98% formic acid were obtained from Ranbaxy Fine Chemicals Ltd. (New Delhi, India), and HPLC-grade water from Thermo Fisher Scientific (Bremen, Germany). Single-use lancets (Avinash Medicals, Bengaluru, India) were used for capillary blood collection and a 350 µL V-bottom collection plate was obtained from Waters Corporation (Milford, CT, USA).

2.2. Sample Preparation for Intact Mass Spectrometric Analysis

Venous samples collected in K₂ EDTA tubes were prepared by mixing 5 µL of blood with 995 µL of HPLC-grade water. The mixture was vortexed and centrifuged at 10,000 rpm (~9200× g) for 5 min. From the resulting supernatant, 20 µL was combined with 180 µL of MS-compatible solvent (50% acetonitrile containing 0.2% formic acid) following a previously described procedure [38].

Capillary blood collected via finger prick was spotted onto DBS cards and DBM BioSampler devices. DBS cards were air-dried at room temperature for 3–4 h before storage and transportation to the laboratory. Circular disks (3.2 mm, corresponding to approximately 3.0 µL blood) were punched from the DBS cards and immersed in 250 µL of HPLC-grade water. Samples were sonicated for 45 min to extract Hb. DBM BioSampler devices containing approximately 30 µL of blood were processed similarly, with each device immersed in 2.5 mL of water and sonicated for 45 min. The extracted supernatants from DBS and DBM samples were subsequently processed in the same manner as venous blood samples for MS analysis [39].

2.3. Instrumentation

2.3.1. Ultra-High-Performance Liquid Chromatography

Samples were delivered to the MS using an Acquity H-Class Plus UHPLC system (Waters, Milford, CT, USA) coupled to the MS, bypassing chromatographic separation. The mobile phase comprised solvent A (0.2% formic acid in water) and solvent B (0.2% formic acid in acetonitrile), which was continuously delivered at a 50:50 ratio. The flow rate profile was programmed as follows: 0.3 mL/min initially; decreased to 0.035 mL/min between 0.15 and 0.40 min; rose to 0.6 mL/min from 0.50 to 0.75 min; and finally returned to 0.3 mL/min till 1.0 min.

2.3.2. Flow Injection Analysis–Triple Quadrupole Mass Spectrometry (FIA-TQMS)

Mass spectrometric detection was performed using a Xevo TQMS (Waters, Milford, CT, USA) operated in positive ion mode using electrospray ionization (ESI). MS settings included a capillary voltage of 2.5 kV, cone voltage of 30 V, source temperature of 150 °C, desolvation temperature of 500 °C, and collision energy of 3 eV. A low collision energy of 3 eV is applied to promote gentle declustering and stable ion transmission while remaining below the threshold for protein fragmentation. It also maintains charge-state envelopes required for deconvolution. Data was acquired at unit resolution with ± 0.75 Da mass tolerance using MassLynx v4.1 software (Waters, Milford, CT, USA).

2.3.3. Complete Blood Count

CBC parameters were analyzed using a Sysmex XN1000 automatic analyzer (Sysmex Corporation, Kobe, Japan). The analyzer operates on the principles of fluorescence flow cytometry, hydrodynamic focusing, and impedance-based cell counting, enabling precise quantification of red blood cell (RBC) indices, white blood cell differentials, Hb concentration, and platelet parameters. Cyanide-free sodium lauryl sulfate Hb methodology was employed for Hb measurement, while advanced optical and impedance technologies were used for cell classification and sizing. CBC analysis requires no manual sample preparation beyond proper collection of WB in K₂ EDTA tubes, gentle mixing, and routine pre-analytical checks for issues like clots, hemolysis, and insufficient volume. All dilution and cell processing steps are performed automatically by the analyzer. The system was calibrated according to the manufacturer’s recommendations, and internal quality control was performed daily using commercial control materials to ensure analytical accuracy and precision, in line with established laboratory standards. Measured CBC parameters included white blood cell count (WBC), red blood cell count (RBC), hemoglobin (Hb), hematocrit (Hct), mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), red cell distribution width (RDW), platelet count (PLT), and differential leukocyte counts. For detection of thalassemia indices, only six parameters were utilized—Hb, RBC count, MCV, MCH, MCHC and RDW.

2.3.4. Capillary Electrophoresis

CE was conducted using a CAPILLARYS 3 Octa system (Sebia, Lisses, France) in Hb (E) mode. The Hb (E) kit employs a buffer optimized for charge-based separation of Hb variants, generating an electropherogram with migration patterns distributed across 15 predefined zones. WB collected in K₂ EDTA tubes was automatically diluted with a hemolyzing solution and introduced at the anodic end of a 25 µm inner-diameter silica capillary. Electrophoresis was performed at 9800 V and 45 µA per capillary under controlled temperature conditions (34 °C), with Hb detected at 415 nm at the cathodic end. Separation was achieved in an alkaline buffer (pH 9.4) based on differences in electrophoretic mobility. The system was calibrated using manufacturer-supplied pooled human blood calibrators, and daily quality control checks were performed. Hb migration times (0–300 s) were recorded and analyzed using Phoresis software v9.3.0.

2.3.5. Sanger Sequencing

Molecular confirmation of β-thal was carried out by Sanger sequencing. PCR-amplified fragments of the β-globin gene were analyzed using the BigDye Terminator v3.1 Cycle Sequencing Kit on Applied Biosystems 3730xl DNA Analyzer (Thermo Fisher Scientific, Waltham, MA, USA). Sequence analysis was carried out using Geneious software v2022.0.1, with the HBB reference sequence as NG_059281.1

2.4. Mass Spectrometry Data

2.4.1. Data Acquisition

Intact globin chain measurements were obtained using a two-channel mass spectrometric acquisition strategy with FIA as a sample introduction method. In this approach, samples were injected into a continuously flowing mobile phase and transferred directly to MS, enabling rapid analysis without chromatographic separation of globin chains. The first channel monitored three charge states each for α, β, δ, ^Gγ, ^Aγ and Hb variant β^S globin chains in multiple reaction monitoring (MRM) mode across the m/z 900–1100 range, chosen to minimize non-overlap of ions from different globin chains. Although the acquisition was implemented using the MRM function of the instrument, no precursor fragmentation was induced in q2. The same precursor m/z was, therefore, monitored in both Q1 and Q3, effectively constituting a non-fragmenting Q1/Q3 mass-filtered measurement rather than a classical precursor–product ion transition. The charge states monitored in this channel were α (16⁺–14⁺: 946.4, 1009.4, 1081.4 m/z), β (17⁺–15⁺: 934.3, 992.6, 1056.8 m/z), δ (17⁺–15⁺: 937.7, 996.3, 1062.6 m/z), ^Gγ (17⁺–15⁺: 942.0, 1000.8, 1067.7 m/z), ^Aγ (17⁺–15⁺: 942.8, 1001.6, 1068.7 m/z) and β^S (17⁺–15⁺: 932.4, 990.6, 1056.6 m/z). For calculating the ratios, only β and δ chains were considered. The second channel employed full-scan acquisition (m/z 650–1200) for deconvolution-based identification of Hb variants. The selected range resulted in the generation of high-intensity mass spectral peaks consistently under the experimental conditions. Deconvolution was carried out using the five most intense peaks within a dual-channel workflow adapted from previously validated methodologies [38].

2.4.2. Data Analysis

For targeted relative quantitation, extracted ion chromatograms (XICs) were generated for each monitored charge state. The intensities of the peak areas of the three XICs corresponding to a given globin chain were summed to obtain a composite signal, which was used to calculate globin chain ratios (e.g., δ/β). δ/β ratios were calculated and are expressed as percentages multiplying by 100. Extraction of peak areas and quantitative data processing were performed using the IonLynx application module.

For variant identification, chromatographic profiles were integrated to obtain the mass spectral peaks. Raw m/z data were internally calibrated with α-globin m/z reference peaks to improve mass accuracy. Five intense peaks from the calibrated spectra were deconvoluted using MaxEnt 1 (Waters, Milford, CT, USA) to derive the neutral masses of intact globin chains and potential variants. The deconvolution parameters included a Gaussian model (resolution 0.75 Da), 80% intensity threshold, and iterative convergence within the 14,000–18,000 Da mass range. The methodology is detailed in Figure 1.

2.5. Statistical Analysis

All statistical analyses were performed using SPSS software v27 (IBM Corp., Rochester, MN, USA). Data variability across sampling techniques was assessed using box-and-whisker plots. The Mann–Whitney U test, a non-parametric test, was applied to evaluate the statistical significance of differences between sampling methods, with p < 0.05 considered statistically significant. Diagnostic performance was assessed using receiver operating characteristic (ROC) curve analysis. 95% confidence intervals (CIs) for AUC values were derived from nonparametric ROC analysis. The Youden index (J = Sensitivity + Specificity − 1) was determined to identify the optimal cut-off value between sensitivity and specificity and to evaluate the clinical performance of the method.

2.6. AI Assistance

The authors used ChatGPT-5 (OpenAI) to assist in statistical interpretations. All scientific content, data analysis, and interpretations were independently developed and verified by the authors, who take full responsibility for the accuracy and integrity of the work.

3. Results

3.1. Detection of β-Thal Carriers Using MS

β-thal carrier detection in this study was based on the relative quantitation of β and δ globin chains. To exclude the presence of globin chain variants that could confound ratio-based interpretation, full scan m/z spectra were deconvoluted to determine the intact masses of individual globin chains. A representative deconvoluted spectrum demonstrating the absence of variants is shown in Figure 1f (Section 2.4.2). Due to the low abundance of δ and γ globin chains in adult blood, their intact masses were not detectable in the deconvoluted spectra acquired on the TQMS platform. The relative quantitation of β^S compared with β was low and, hence, the presence of the variant was ruled out. No structural variants that deviate from the wild type were detected in any of the samples under the experimental conditions. An elevated δ/β (%) was observed in β-thal carriers compared with normal samples and showed concordance with thalassemia indices derived from CBC, as well as increased HbA2 (%) obtained from CE. A comparative analysis of representative normal and β-thal carrier samples is presented in Figure 2.

3.2. Cascade Screening

The participants represented three generations, with age groups of 61–80, 31–60, and 13–30 years. Participant 7 represented generation I, though they belonged to the age group between 31 and 60 yrs. Females represented 59% and males were 41%. Since Sample (S) number S4, S5 were normal, their descendants were not tested for β-thal. Samples of the descendants of S6, S7 were not collected, since they were not residing in India during the conduct of the experiment. CBC analysis of six participants (S12–S15, S18) and S16 was performed in the USA and New Zealand respectively. All other samples were processed in different labs in Bengaluru, India. The available CBC and CE data of the participants, along with mass spectrometry information from various sampling methods—DBS, DBM, WB—is provided in

Table 1. Sample details of participants. Hb—hemoglobin, RBC—red blood cell, MCV—mean corpuscular volume, MCH—mean corpuscular hemoglobin, MCHC—mean corpuscular hemoglobin concentration, RDW—red blood cell distribution width, HbA0—adult hemoglobin α²β², HbA2—hemoglobin α²δ², DBS—dried blood spot, DBM—dried blood matrix, WB—whole blood, δ—delta globin chain, β—beta globin chain. Complete blood count, Capillary Electrophoresis and WB were performed on venous blood. NA—Not available. * DBS and DBM values are an average of triplicates. ** denotes β-thal carriers. Sample 18 was classified based on prior CBC and molecular results.

Sample No.	Demographics		Complete Blood Count							Capillary Electrophoresis		Mass Spectrometry δ/β (%)
	Sex at Birth	Age (Years)	Hb (g/dL)	RBC Count (Milli/cmm)	MCV (fl)	MCH (pg)	MCHC (g/dL)	RDW (CV%)	Thal Indices	HbA0	HbA2	DBS *	DBM *	WB
1 **	F	61–80	10.1	5.47	67.8	18.5	27.2	18.2	Yes	94.2	5.3	22.15	18.63	16.51
2 **	M	61–80	9.7	4.89	68.5	19.8	29.0	16.9	Yes	96.0	4.0	19.62	18.64	18.18
3	F	61–80	12.8	4.53	90.1	28.3	31.4	13.5	No	97.5	2.5	19.32	17.74	15.46
4	M	61–80	12.7	4.24	92.0	30.0	32.6	13.7	No	97.5	2.5	17.14	15.60	16.49
5	F	61–80	11.2	4.20	89.0	26.7	29.9	13.9	No	97.6	2.4	16.41	18.25	16.78
6 **	M	61–80	9.9	5.14	66.5	19.3	28.9	17.8	Yes	95.6	4.4	19.79	17.03	19.63
7	F	31–60	10.8	4.12	86.4	26.2	30.3	13.7	No	97.7	2.3	17.95	16.29	13.86
8	F	31–60	14.1	4.82	91.3	29.3	32.0	13.2	No	97.8	2.2	17.09	17.03	14.07
9 **	M	31–60	11.0	6.02	68.1	18.3	26.8	17.9	Yes	94.8	4.8	21.52	22.90	18.41
10	F	31–60	12.8	4.77	83.2	26.8	32.2	13.8	No	-	-	17.41	17.30	-
11	M	31–60	15.3	5.56	84.2	27.5	32.6	14.5	No	-	-	18.09	19.25	-
12	F	31–60	12.4	4.64	84.5	26.7	31.6	14.1	No	-	-	-	-	-
13	M	31–60	13.5	4.60	90.2	29.3	32.5	12.8	No	-	-	17.66	14.47	-
14	F	31–60	12.7	4.39	90.2	28.9	32.1	13.6	No	-	-	-	-	-
15	M	31–60	15.8	5.42	85.0	29.2	34.2	12.1	No	-	-	-	-	-
16 **	M	31–60	13.2	7.03	63.0	18.8	NA	NA	Yes	-	-	18.83	-	-
17	F	31–60	13.2	4.06	99.5	32.3	32.4	NA	No	-	-	15.48	-	-
18 **	F	31–60	8.5	3.96	66.9	21.3	31.9	18.5	Yes	-	-	-	-	-
19 **	F	13–30	10.4	5.27	62.6	19.7	31.5	15.4	Yes	95.9	4.1	19.32	20.23	12.47
20	F	13–30	11.3	4.02	87.0	28.1	32.2	12.6	No	97.5	2.5	17.00	17.46	10.20
21	M	13–30	NA	NA	NA	NA	NA	NA	NA	-	-	17.61	16.56	-
22	F	13–30	NA	NA	NA	NA	NA	NA	NA	-	-	18.12	17.24	-

. Seven participants showed thalassemia indices, including S1, S2, S6 (generation I), S9, S16, S18 (generation II) and S19 (generation III), with HbA2 elevated (>4.0). S16 declined to provide venous blood, and, hence, CE and MS (WB) were not performed. Due to the unavailability of a DBM device, only DBS samples were collected for S16 and S17. β-thal carriers showed higher mean values than normal individuals across all sample types. In DBS samples, the mean value in carriers was 20.54 (18.83–22.15) compared with 17.61 (15.48–19.32) in normal individuals. In DBM samples, carriers demonstrated a mean of 19.49 (17.03–22.90), whereas normal individuals showed a mean of 16.93 (14.47–19.25). Similarly, in WB, the mean δ/β (%) value was higher in carriers at 17.04 (12.47–19.63) compared to 14.81 (10.20–16.78) in normal individuals.

3.3. Pedigree Chart

A pedigree chart highlighting β-thal carriers across three generations is presented in Figure 3. Micro sampling approaches coupled with MS identified six positive cases. In addition, one participant (S18) was confirmed based on prior CBC and molecular results, corresponding to 31.8% (7/22) of β-thal carriers in the family. The carrier status reduced to 20.6–23.5% if all members (n = 34) were included. A range is provided to include the possibility of β-thal carriers among the descendants of S18.

3.4. Comparison of Micro Sampling Techniques

To examine the distribution of MS-derived signal intensities, box-and-whisker plot analysis (Figure 4) was performed for each sampling technique. The spread of values for DBS and DBM (capillary-derived samples) did not differ significantly (p > 0.05), indicating a comparable analytical performance. In contrast, both DBS and DBM differed significantly from venous whole blood (p < 0.05), suggesting matrix-dependent variation in ion response or analyte recovery between capillary and venous sampling.

3.5. Diagnostic Performance Across Sample Types

Receiver Operating Characteristic (ROC) curve analysis was performed to evaluate the diagnostic accuracy of mass spectrometry-based assessment of β-thal using capillary blood (DBS, DBM) and venous WB with CE as a reference standard.

3.5.1. Dried Blood Spot (DBS)—Capillary

ROC curve analysis of mass spectrometry results from DBS samples showed an area under the curve (AUC) of 0.98, reflecting excellent diagnostic accuracy in discriminating against β-thal-positive from -negative cases when compared with CE. The optimal cutoff value (~18.6) yielded a Youden index of 0.83, corresponding to 100% sensitivity and 83% specificity. Although a higher threshold (19.47) achieved complete specificity, it reduced sensitivity to 80%, thereby lowering the overall diagnostic efficiency. Hence, the cutoff of 18.63 offered the best balance between false-positive and false-negative classifications. These findings highlight the robustness of DBS as a minimally invasive, capillary-based sampling technique suitable for large-scale or cascade screening applications.

3.5.2. Dried Blood Matrix (DBM)—Capillary

For DBM samples, the ROC curve analysis yielded an AUC of 0.88, consistent with a good diagnostic performance relative to CE. The optimal threshold (~18.4) produced a Youden index of 0.8, with sensitivity of 80% and specificity of 100%. These results demonstrate that DBM performs comparably to DBS in identifying β-thal carriers, supporting its suitability as an alternative micro sampling format, particularly in settings where DBS preparation or transport may be challenging.

3.5.3. Whole Blood (WB)—Venous

ROC curve analysis for venous whole blood demonstrated an AUC of 0.80 ± 0.16 (p = 0.10, 95% CI, 0.49–1.00), indicating good, though not statistically significant, discriminating ability versus CE. The optimal cutoff (~16.5) corresponded to a Youden index of 0.63, with a sensitivity of 80% and specificity of 83%. The wide confidence interval likely reflects the small sample size (n = 11). Nonetheless, these data suggest that WB mass spectrometric analysis remains a reliable approach for confirmatory testing. The statistical parameters highlighting the diagnostic accuracy of various formats are provided in Figure 5.

3.6. Clinical Performance Across Sample Types

When compared with CE, DBS achieved the highest sensitivity (100%), identifying all CE-positive samples correctly, though one false positive reduced its specificity to 83%. DBM showed perfect specificity (100%) but a slightly reduced sensitivity (80%) due to one missed case. Whole blood showed a lower overall accuracy (82%), with one false-positive and one false-negative result. A comparative analysis of the sampling methods is shown in

Table 2. Clinical validation of samples analyzed using DBS, DBM, and WB sampling methods with CE as the reference method. Diagnostic performance is expressed in terms of sensitivity, specificity and overall accuracy. CE—capillary electrophoresis, DBS—dried blood spot, DBM—dried blood matrix, WB—whole blood, +ve Positive, −ve Negative.

Sample Type	CE +ve	CE −ve	Sensitivity (%)	Specificity (%)	Accuracy (%)
DBS +ve	5	1	100	83	91
DBS −ve	0	5
DBM +ve	4	0	80	100	91
DBM −ve	1	6
WB +ve	4	1	80	83	82
WB −ve	1	5

.

Overall, both DBS and DBM demonstrated high agreement (91%) with CE, supporting their reliability as micro sampling alternatives for β-thal screening, while WB showed a marginally lower performance.

4. Discussion

The analysis of globin chains using flow injection analysis–mass spectrometry (FIA-MS) in a high-throughput manner is well suited for both population and carrier screening. Compared to population-based screening, which is resource-intensive and logistically demanding, carrier screening offers a more targeted strategy that yields higher diagnostic accuracy. Carrier screening of β-thal can be implemented through three primary approaches: (i) cascade screening of the relatives of affected individuals, (ii) antenatal or preconception screening of couples, and (iii) population screening of all individuals of reproductive age [40].

In the present study, the index case (S19) was initially diagnosed as a β-thal carrier using gel electrophoresis. Two decades later, her parents sought to trace the carrier status within the extended family. The family members were distributed across different Indian states (Tamil Nadu, Karnataka, Andhra) and countries (USA, Dubai, New Zealand). In the absence of direct access to all relatives during the study period, only available CBC parameters were included. Capillary blood collection (finger prick) was employed to overcome challenges associated with venous sampling. Participants readily consented to capillary collection, whereas one member (S16) declined to provide venous sampling. The authors personally conducted sample collection at a common residential site.

Among micro sampling techniques, DBM collection was observed to be less cumbersome and did not require drying, unlike DBS, highlighting its potential as a practical alternative. With minimum training, capillary self-sampling could be implemented in community-based or remote settings without supervision.

Ratios of globin chains such as α/β, δ/β (%), δ/α (in the presence of variants), γ/α, and γ/β serve as reliable biomarkers for identifying and differentiating β-thal types, including β-thal, δβ thal, and HPFH (Hereditary Persistence Fetal Hemoglobin) [35,36,37]. In this study, none of the participants had a variant (based on full-scan analysis from channel 2), that confounds the quantification of globin chains. While α/β ratio is a reliable biomarker for α-thalassemia, it performs poorly in detecting β-thal carriers due to compensatory globin chain synthesis [36]. In contrast, the δ/β ratio is a more informative biomarker for β-thal, as reduced β globin production leads to compensatory upregulation of δ globin synthesis. Expressing δ/β as a percentage normalizes the relative abundance of δ globin to β globin, enabling direct comparison across samples despite variation in globin concentration. The resulting increase in δ/β (%) directly mirrors the elevation of HbA2 observed by CE, establishing a strong concordance between MS-derived δ/β measurements and CE-based HbA2 quantification.

The micro sampling technique combined with TQMS identified the β-thal trait in 20.6 % of family members. Although the participants currently reside in south India, their ancestral origin traces back to the southern region of Gujarat, from where their forebears migrated approximately 1000 years ago [41]. The detection of carrier status in two out of three siblings of S1 (the grandparent of the index case) underscores the utility of cascade screening in tracing carriers through family networks [Figure 3]. The MS-based findings were concordant with the CBC and CE results [Table 1], which showed characteristic features of β-thal-trait-reduced MCV (<80 fl) and MCH (<25 pg), along with increased HbA2 levels (>3.5%) [42]. The family carried HBB: c.92 + 5 G > C, one of the most prevalent β-thal mutations in India [43], which affects RNA processing and results in a β⁰ or β⁺ phenotype [44]. The prevalence of the mutation varies from 44.8% in the north to 71.4% in the Eastern India [43]. Although carrier detection has no direct clinical impact on individual health, its identification is crucial for genetic counseling and prevention of β-thal major births.

This study compared the performance of three sampling techniques—capillary-based micro sampling (DBS, DBM) and venous WB—for β-thal carrier detection. DBS and DBM produced comparable results, with no statistically significant difference between them, confirming their suitability for community and outreach screening programs. However, both micro sampling techniques showed a statistically significant difference (p < 0.05) when compared with venous WB (Figure 4). A higher δ/β (%) ratio was observed in capillary samples relative to venous blood, likely due to the higher Hb concentration in capillary samples, as reported by Becker et al. [45].

All micro sampling specimens were analyzed in a single batch, whereas venous WB samples were processed on separate days, depending on availability. Such differences, combined with the small sample size (reflected by wide 95% confidence intervals, Figure 5), may have contributed to the minor diagnostic variation between sampling types. The observed variation was statistically significant but not clinically significant, consistent with previous reports for analytes such as sodium, potassium, chloride, phosphate, creatinine and total protein [46]. These differences likely arise from matrix effects, including blood composition, analyte-specific bias, and hematocrit variability, supporting the need for technique-specific reference ranges for different sampling methods [11,47,48].

DBS-MS had 100% sensitivity and DBM-MS had 100% specificity (Table 2), demonstrating the feasibility of using micro sampling techniques for population-level carrier screening of β-thal. DBS-MS is already established for detecting in-born errors of metabolism in neonates [49] and has been extended to identify clinically significant hemoglobin variants (HbS, HbC, HbE, HbD-Punjab, Hb O-Arab) and β-thal major [50]. The current findings indicate that DBS-MS and DBM-MS can also be applied to adult carrier detection, where β globin chain quantification is diagnostically relevant. These micro sampling approaches in combination with MS techniques (encompassing both LC-MS/MS and Gas Chromatography–MS techniques) are particularly suited for large-scale, multi-biomarker population studies [51,52]. Among these, DBM offers practical advantages—ease of use, no drying requirement, and compatibility with self-sampling in remote areas—facilitating centralized analysis in MS-equipped laboratories.

A few limitations should be acknowledged. The small sample size, geographical dispersion of participants and collection at different times may have introduced pre-analytical variability, particularly for venous WB samples. Larger, standardized studies are needed to validate these observations.

In summary, cascade screening using micro sampling and mass spectrometry represents a sensitive, scalable and practical strategy for β-thal carrier detection in adults. Integration of these technologies into existing clinical and public health frameworks could significantly strengthen screening capacity, particularly in resource-limited or geographically diverse populations.

5. Conclusions

In this preliminary study, the application of FIA-TQMS, in combination with conventional CBC parameters, using micro sampling techniques, demonstrated the potential for detecting β-thal carriers in adults, thereby broadening the diagnostic scope of hemoglobinopathy evaluation in clinical laboratories. The identification of β-thal carriers within high-risk families suggests a targeted and practical alternative to large-scale population screening. However, the current data are insufficient to draw definite conclusions regarding its clinical performance characteristics, including sensitivity, specificity, and overall accuracy. Additionally, the influence of preanalytical variables—ranging from blood collection to sample handling and processing—on carrier detection warrants systematic investigation. Although DBS and DBM approaches showed a comparable performance, quantitative differences relative to venous blood highlight the need for sampling-technique-specific reference intervals. Despite limitations in sample size and collection conditions, these findings support the feasibility of integrating MS-based micro sampling workflows into cascade screening strategies. Such an approach may enhance accessibility and early identification of β-thal carriers; however, comprehensive clinical validation in larger and genetically diverse cohorts is required before FIA-TQMS-based strategies can be recommended for population-level or preventive β-thal screening.