Analytical and Clinical Evaluation of a Camel Polyclonal Antibody-Based IRMA for Serum Thyroglobulin Measurement Compared with ECLIA in Thyroid Cancer Patients

INTRODUCTION

Differentiated thyroid cancer (DTC), comprising papillary and follicular carcinomas, accounts for over 90% of thyroid malignancies. Its rising global incidence, particularly among younger individuals and females, highlights the need for reliable strategies for long-term surveillance. Standard management involves total or near-total thyroidectomy followed by radioactive iodine ablation. Post-treatment monitoring is critical for early detection of recurrence or metastasis, where timely intervention improves outcomes. Serum thyroglobulin (TG), a 660 kilodalton (KD) glycoprotein produced exclusively by thyroid follicular cells, serves as a highly specific biomarker for residual or recurrent disease.^[1]

Following successful thyroid ablation, TG levels are expected to be undetectable or very low. Detectable TG during follow-up may indicate persistent or recurrent disease. However, the clinical utility of TG measurement is limited by interference from anti-thyroglobulin antibodies (TGAB), present in approximately 20% of DTC patients.^[2] These antibodies can mask epitopes or alter assay binding, resulting in falsely low TG values and potential misinterpretation of disease status.^[3]

Multiple assay platforms have been developed for TG estimation, differing in analytical sensitivity, specificity, and susceptibility to TGAB interference. Electrochemiluminescence immunoassay (ECLIA) systems are widely used due to their high sensitivity, automation, and rapid turnaround time. However, they require significant financial investment, infrastructure, and maintenance, limiting their accessibility in resource-constrained settings. Moreover, despite international standardisation using a certified reference material (CRM-457), inter-assay variability persists, and TG results from different platforms remain non-interchangeable.^[4]

Such variability has important clinical implications, as inconsistencies in TG measurements may lead to unnecessary investigations or inappropriate treatment decisions. Maintaining assay consistency for individual patients is therefore essential but often difficult in real-world settings, particularly in public healthcare systems. This has prompted interest in developing cost-effective and locally adaptable alternatives.

Immunoradiometric assays (IRMA), especially those developed in-house, offer a viable alternative due to their high specificity and potential for cost reduction through bulk reagent preparation. These assays can be tailored to local laboratory requirements, making them suitable for high-volume and resource-limited settings.^[5] However, comparative data evaluating the performance of in-house IRMA systems against widely used commercial platforms remain limited.

The in-house IRMA evaluated in this study was previously validated against a commercially available IRMA and demonstrated satisfactory analytical performance.^[5,6] With the increasing adoption of ECLIA platforms in routine clinical practice, a cross-platform comparison has become necessary to assess agreement and clinical reliability. The present study, therefore, evaluates the analytical and clinical performance of the in-house IRMA in comparison with ECLIA in thyroid cancer patients.

Notably, this IRMA employs camel-derived polyclonal anti-thyroglobulin antibodies as the capture reagent in combination with anti-TG monoclonal antibodies, a strategy rarely described in TG immunoassay methodologies for routine clinical use. Apart from prior work from our group, evidence on their application remains limited.^[5,6] This study provides a comparative evaluation of this indigenously developed assay with a standard ECLIA, offering insights into its potential utility in routine clinical practice.^[7]

MATERIAL AND METHODS

RESULTS

A total of 157 serum samples were analysed, representing a demographically relevant patient group aged 11–79 years, with a male-to-female ratio of approximately 1:2. This skewed distribution aligns with known epidemiological trends, wherein thyroid cancer disproportionately affects females.

The degree of agreement between the two methods was assessed using linear regression. The linear regression model, Y = 0.18X + 4.8, highlights a modest association, indicating limited agreement. The correlation coefficient (r) was calculated to evaluate the statistical relationship between the IRMA and ECLIA results across all tested specimens. Pearson’s correlation analysis revealed a moderate and statistically significant positive correlation between the two methods (r =0.51, n = 157, p <0.001) [Fig 1].

Close

Fig 1: Scatter plot of TG levels by IRMA and ECLIA (r = 0.51, n = 157, p <0.001). The regression equation (y = 0.18x + 4.8) indicates analytical comparability. IRMA: Immunoradiometric assays; ECLIA: Electrochemiluminescence immunoassay; TG: Thyroglobulin.

Export to PPT

While correlation coefficients provide a measure of association, they do not evaluate agreement. To address this, a Bland– Altman plot was employed to assess the extent of bias and limits of agreement between the two methods [Fig 2]. The plot displayed the differences in TG values (IRMA – ECLIA) plotted against the mean of both methods. The overall mean bias was positive, reinforcing the earlier observation that IRMA slightly overestimates TG levels compared to ECLIA, particularly at the lower end of the concentration spectrum.^[9]

Close

Fig 2: Bland–Altman plot comparing serum thyroglobulin levels measured by in-house IRMA and Eleusis TG II (ECLIA), showing a moderate positive correlation between their mean. IRMA: Immunoradiometric assays; ECLIA: Electrochemiluminescence immunoassay; TG: Thyroglobulin.

Export to PPT

Importantly, most data points lie within the 95% limits of agreement, denoted by dotted lines, suggesting that while minor proportional bias exists, the agreement is clinically acceptable for many diagnostic and follow-up scenarios. However, a few outliers were observed, particularly at higher TG concentrations, which could reflect hook effect, TGAB interference, or varying assay response at elevated antigen loads. Such variability underscores the need for method-specific cutoffs and clinical interpretation thresholds.^[10]

Another key point of distinction was the reference range strategy adopted by each method [Table 2]. While ECLIA applies a fixed reference range (3.5–77 ng/ml) regardless of clinical context, the IRMA method incorporates stratified intervals tailored to the patient’s thyroid status and treatment regimen (e.g., intact thyroid vs thyroidectomised patients on/off T4 therapy).

DISCUSSION

Accurate monitoring of serum TG levels is a cornerstone in the follow-up of DTC patients. Given that different assay platforms employ distinct detection principles and exhibit varied analytical sensitivities, it is essential to maintain consistency by using the same method across all serial measurements. This approach reduces variability and ensures consistent interpretation of disease status, especially when monitoring potential recurrence or remission.^[1,11]

Although it is important to explore alternative methods, we must carefully evaluate their effects on patient management. Inter-assay variability among commercial immunoassays remains a challenge, even with standardisation efforts like using CRM-457 as a reference material.^[12] Studies have shown significant differences in TG values across various platforms. These differences can lead to misinterpretation of trends in patient monitoring.^{[8, 13]}

In this study, we evaluated the analytical performance of both TG assay platforms by examining several key parameters. These included the working range, which is the concentration range where the assay maintains acceptable accuracy and precision, and functional sensitivity, defined as the lowest concentration of TG that can be reliably measured with a set precision level, usually a 20% coefficient of variation. We also looked at reference intervals and statistical comparability. Together, these parameters provide a complete assessment of the assay’s accuracy, reliability, and clinical use in measuring serum TG levels, especially for thyroid cancer monitoring. Inter-assay variability continues despite using certified reference materials like CRM-457. These discrepancies mostly arise from differences in calibrator composition, the natural variation of TG, and the use of assay-specific antibody sets with different affinities and specificities toward various TG isoforms. Such variability can greatly impact the interpretation of serial TG measurements, which are crucial for monitoring and assessing the risk of patients with DTC. Current clinical guidelines, including those from the American Thyroid Association (ATA), stress the need to use the same assay consistently for ongoing monitoring. When changing assays is unavoidable, it is advisable to re-establish baseline values and interpret trends carefully to ensure appropriate clinical decisions. Therefore, assay standardisation, along with careful methodological consistency, is vital to improve the reliability of TG as a tumour marker in managing DTC patients.^[14,15] The Eleusis TG II (ECLIA) method showed a measuring range of 0.04 to 500 ng/ml, with a functional sensitivity of 0.1 ng/ml. In contrast, the in-house IRMA assay worked over a range of 0.1 to 300 ng/ml, with a functional sensitivity of 0.3 ng/ml. While ECLIA demonstrated slightly greater analytical sensitivity (0.04 ng/ml) than IRMA (0.1 ng/ml), both methods effectively measured TG concentrations across clinically relevant ranges.

Cut-off values may vary significantly across different TG assays, underscoring the need for assay-specific definition of cut-off and clinical decision thresholds to ensure accurate interpretation and patient management.^[16] Table 2 highlights an important difference in the reference intervals applied by the two methods, particularly in relation to thyroid stimulating hormone (TSH) status. ECLIA employs a uniform reference range of 3.5–77 ng/mL across all patient groups, irrespective of TSH levels or thyroid status. In contrast, the IRMA assay incorporates condition-specific cutoffs aligned with TSH-suppressed and TSH-stimulated states: undetectable to 30 ng/mL in individuals with intact thyroids, undetectable to 14 ng/mL in thyroidectomised patients off T4 therapy (TSH-stimulated), and undetectable to 3.5 ng/mL in those on T4 therapy (TSH-suppressed).^[17] This distinction is clinically important because TG levels are TSH-dependent: elevated TSH increases TG production, while TSH suppression lowers it. Ignoring TSH status may therefore lead to misinterpretation. Incorporating TSH-based cutoffs enhances the clinical relevance of results, particularly in post-operative surveillance. Despite the observed variability in absolute values, both methods demonstrated directional consistency, a crucial factor for longitudinal monitoring. However, the limited interchangeability implies that using both assays interchangeably during patient follow-up may introduce interpretation errors. Therefore, continued use of a single, validated method is recommended for each patient over time.

Pearson’s correlation analysis showed a moderate and statistically significant positive correlation between the methods (r = 0.51, n = 157, p <0.001), suggesting both platforms generally followed the same trend. However, the moderate correlation coefficient may reflect differences in assay architecture, antibody specificity, calibration procedures, and epitope recognition characteristics between the two assay systems. Additionally, circulating TG heterogeneity and possible TGAB-related interference may have contributed to inter-assay variability. The linear regression equation y =0.18x + 4.8 was obtained, indicating a modest relationship. The low slope (0.18) suggested that increases in ECLIA-derived TG values were associated with smaller increases in IRMA results, while the positive intercept (4.8) reflected a tendency of IRMA to report slightly higher values at lower TG levels. However, as the study was not specifically designed for low-range analytical performance evaluation, definitive conclusions regarding superior sensitivity at lower TG levels cannot be drawn.

The scatter plot [Fig 1] demonstrated this trend visually. A concentration of values appeared below 100 ng/ml, where the agreement between the two methods was relatively tighter. As values increased, the spread between methods widened, with some clear outliers near the assay limits, particularly at 500 ng/ml on ECLIA.

To further evaluate agreement, a Bland–Altman plot was generated [Fig 2]. This plot assessed the difference between the two methods against their mean values. The visual showed that while most data points fell within the 95% limits of agreement (dotted lines), several points deviated significantly at both low and high ends. These differences may stem from hook effect susceptibility, calibration discrepancies, or the presence of TG autoantibodies.

Together, the scatter and Bland–Altman plots confirmed that although the assays showed directional consistency, they were not directly interchangeable. The IRMA method showed a slight bias toward higher baseline values, and method-specific interpretation remained necessary. Clinical follow-up using only one consistent method is therefore recommended to ensure longitudinal accuracy and reduce analytical variability.^[18]

CONCLUSION

The findings of this comparative study indicate a moderate yet statistically significant correlation between the in-house IRMA and the Eleusis ECLIA methods. However, the observed variation in absolute values reinforces the recommendation that TG results from different assays should not be interchanged without caution.

The indigenous IRMA kit has been in continuous use for over 12 years at our centre, with over 50,000 samples analysed, demonstrating its robustness and applicability in routine clinical workflows. One of the key advantages of the IRMA platform lies in its cost-effectiveness. Bulk preparation of reagents significantly lowers production costs without compromising assay performance. This not only ensures sustainability within high-volume testing environments but also makes IRMA a practical solution for resource-limited healthcare settings, where access to advanced non-isotopic platforms like ECLIA may be constrained.

A key limitation of this study is the absence of prospective follow-up data, which would provide additional insight into the long-term clinical utility of the assay. Nevertheless, the indigenously developed camel polyclonal anti-TG IRMA has been used routinely in patients for the past 12 years, supporting its practical applicability and relevance in monitoring differentiated thyroid cancer. Additionally, the study compared a radio isotopic IRMA method with a non-isotopic ECLIA platform, two fundamentally different technologies with inherent differences in detection principles. These methodological disparities may contribute to variations in measured values and are difficult to fully standardise.

Taken together, the study supports the continued use and further expansion of indigenously developed, affordable assay systems like IRMA for thyroid cancer monitoring, especially in settings where affordability, accessibility, and long-term consistency are paramount.