Introduction

Thyroid cancer is the most frequently diagnosed malignancy among both men and women in Korea, according to the Korea Cancer Incidence Database (1999–2021) [1]. In 2020, thyroid cancer exhibited the highest prevalence among all cancers, with an overall rate of 953.6 per 100,000 population. The sex prevalence ranged from 357.5 per 100,000 in men to 1,546.6 per 100,000 in women [2]. Clinically, patients with thyroid cancer often present with thyroid nodules, with or without an associated goiter [3]. Accurate identification of nodules indicative of malignancy is crucial, as differentiated thyroid cancer (DTC) represents the most prevalent histological subtype [4].

The standard treatment for DTC typically involves thyroidectomy, with high-risk patients receiving radioactive iodine ablation and thyroid-stimulating hormone (TSH) suppression therapy [5]. Serum thyroglobulin (Tg), synthesized exclusively by normal and well-differentiated thyroid follicular cells, is the most reliable and widely used tumor marker exhibiting high tissue specificity and is highly used for posttreatment surveillance of DTC [6]. Detectable Tg levels following total thyroidectomy and radioactive iodine ablation indicate persistent, recurrent, or metastatic disease. Consequently, Tg measurement, particularly when performed using highly sensitive immunometric assays alongside anti-Tg antibody testing, has become an essential component of the long-term monitoring of patients with DTC [7-9].

Tg is a high-molecular-weight glycoprotein (approximately 660 kDa) synthesized by thyroid follicular cells and stored in the follicular lumen, where it plays a central role in the synthesis of thyroid hormones [10]. Serum Tg measurement is essential for follow-up and long-term management of patients with DTC [11]. Elevated Tg levels may indicate residual or recurrent disease, although they can also increase in response to thyroid injuries, such as surgery, biopsy, or radioactive iodine therapy [12,13]. Tg functions as a precursor for the synthesis and intrafollicular storage of thyroxine (T4) and triiodothyronine (T3) [14]. Although low Tg levels are present in the serum of healthy individuals, their concentrations may increase in the presence of thyroid nodules, tissue damage, or increased TSH receptor activity [15]. Therefore, serum Tg serves as an important biomarker in various clinical contexts and should be interpreted considering the underlying physiological and pathological conditions of the patient.

The Siemens Atellica IM Tg assay, commercially launched in Korea in July 2024, is an in vitro diagnostic test developed for the quantitative measurement of Tg levels in human serum, and is intended for monitoring thyroid function before and after thyroid cancer treatment. However, their analytical performances have not yet been fully evaluated. This study aimed to assess the performance of the Atellica IM Tg assay by examining key analytical parameters and comparing the results with those of the established Roche Cobas Elecsys Tg II assay.

Methods

Ethical statements: The study protocol was approved by the Institutional Review Board (IRB) of the Kosin University Gospel Hospital (KUGH) (IRB No. KUGH 2023-03-007) and conducted in accordance with the principles outlined in the Declaration of Helsinki. The requirement for informed consent was waived because the study used leftover serum specimens collected during routine clinical testing.

Analytical performance parameters, including precision, linearity, limit of blank (LoB), limit of detection (LoD), limit of quantitation (LoQ), and reference ranges, were assessed. A comparative analysis was performed using the Roche Cobas Elecsys Tg II assay. Serum samples were used to evaluate the correlation, categorical concordance, and agreement between the two assays across clinically relevant Tg concentration ranges.

1. Precision evaluation

To assess precision, the coefficient of variation (CV) was evaluated using commercially available quality control (QC) materials, specifically Bio-Rad Liquichek Tumor Marker Controls (Bio-Rad Laboratories, Inc.). Three QC levels were used with target concentrations of 4.18 ng/mL (low), 44.02 ng/mL (intermediate), and 105.64 ng/mL (high). Each QC level was tested over a 20-day period with two runs per day and two replicates per run, following the guidelines outlined in the Clinical and Laboratory Standards Institute (CLSI) EP15-A3 protocol [16-18]. The repeatability and within-laboratory precision were determined by calculating the CV values and comparing them with the manufacturer’s specifications provided in the product insert.

2. Linearity validation

Linearity was evaluated according to the CLSI EP06-Ed2 guidelines [19]. A high-concentration patient serum sample was serially diluted with a low-level matrix at ratios of 4:0, 3:1, 2:2, 1:3, and 0:4, resulting in five different concentration levels. Each dilution was tested in duplicate. Data analysis was performed following the procedures described in the CLSI EP06-Ed2. The allowable deviation from linearity (ADL) was set at 10% in this study, and deviations were calculated for each concentration level. Linearity was considered acceptable if the observed deviations were within the predefined ADL threshold.

3. Verification of LoB, LoD, and LoQ

The LoB, LoD, and LoQ were verified according to the CLSI EP12-A2 guidelines [20]. For LoB and LoD verification, blank and low-concentration patient samples were tested 20 times. To verify the LoQ, five patient samples with concentrations near the manufacturer’s claimed LoQ were analyzed, with each sample tested in 10 replicates. The LoQ was confirmed if the results met the predefined criteria for the total allowable error.

4. Verification of reference intervals

The reference interval for Tg levels specified by the manufacturer (1.82–111 ng/mL) was verified using serum samples from an apparently healthy cohort. Residual specimens collected during routine health examinations unrelated to this study were used for analysis. The inclusion criteria for the healthy cohort were as follows: age between 18 and 50 years, absence of chronic medical conditions, no current medication use, and no recent history of hospitalization. A total of 20 qualifying samples were obtained from the KUGH. Subsequently, the samples were evaluated based on the CLSI EP28-A3 guidelines [21]. This verification assessed the applicability of the manufacturer’s reference interval to the local population in the present study.

5. Comparative evaluation of Tg assays

We collected 681 anonymized patient samples from KUGH for the comparative evaluation of Tg assay results. The samples were transferred into serum separator tubes, divided into two aliquots, and analyzed independently. One aliquot was tested in a routine laboratory using the Cobas Elecsys Tg II assay, and the other was analyzed using the Atellica IM Tg assay. The Tg values obtained from the two analyzers were categorized into three clinical decision ranges: low (<3.5 ng/mL for Cobas Elecsys Tg II and <1.82 ng/mL for Atellica IM Tg), normal (3.5–77 ng/mL for Cobas Elecsys and 1.82–111 ng/mL for Atellica IM), and high (>77 ng/mL for Cobas Elecsys and >111 ng/mL for Atellica IM). Methods comparison was performed using Passing-Bablok regression and Bland-Altman analyses. The correlation coefficient (Pearson’s r) between the two assays was calculated. Additionally, the agreement between the two assays was evaluated using a weighted kappa statistic to assess categorical concordance.

6. Statistical analyses

Statistical analyses, including assessments of precision, linearity, Bland-Altman plots, Passing-Bablok regression, and correlation coefficients, were performed using Analyze-it software version 5.90 (Analyze-it Software Ltd.). Weighted kappa statistics were calculated using MedCalc 23.3.3 version software (MedCalc Software Ltd.). Statistical significance was set at p<0.05.

Results

1. Precision assessment of Atellica IM Tg

For the Atellica IM Tg assay, the repeatability and within-laboratory CV were 3.3% and 3.3% for the low-level QC material (4.18 ng/mL), 1.8% and 2.5% for the intermediate level (44.02 ng/mL), and 1.4% and 2.6% for the high level (105.64 ng/mL), respectively (Table 1). All CV values were within the acceptable limits set by the manufacturer, demonstrating satisfactory precision throughout the tested concentration range.

2. Detection capability

The mean LoB, LoD, and LoQ values were 0.011, 0.015, and 0.040 ng/mL, respectively, which met the manufacturer’s specifications for LoB (<0.026), LoD (<0.036), and LoQ (<0.050).

3. Linearity validation

Linear regression analysis confirmed the excellent linearity of the assay across a concentration range of 0.05–152.49 ng/mL. All observed deviations from linearity for the prepared dilution levels remained within the predefined allowable limits (Fig. 1), thereby validating the assay’s linear performance throughout the tested range.

4. Reference range verification

Of the 20 tested samples, 18 had Tg concentrations within the manufacturer’s proposed reference interval of 1.82–111 ng/mL. Although the two samples did not fall within this range, the results met the CLSI EP28-A3 acceptance criteria. This supports the adoption of the manufacturer’s reference interval for use in the local population.

5. Comparative analysis of Atellica IM Tg and Cobas Elecsys Tg II assays

For the correlation analysis between the two assays, 104 patient samples with results outside the analytical measurement range (AMR) of either assay were excluded from the study. The remaining 577 patient samples were used exclusively for correlation analysis, which demonstrated excellent agreement (Pearson’s r=0.997) (Fig. 2).

Concordance between the two assays was evaluated using 681 patient samples stratified into three clinical Tg concentration ranges. In the low range, defined as <3.5 ng/mL for the Roche Cobas Elecsys Tg II assay and <1.82 ng/mL for the Siemens Atellica IM Tg assay, the concordance rate was 83% (Table 2). In the normal range (Roche Cobas, 3.5–77 ng/mL; Siemens Atellica, 1.82–111 ng/mL), the concordance increased to 98%. At high concentrations (Roche Cobas, >77 ng/mL; Siemens Atellica, >111 ng/mL), the concordance decreased to 65%. The overall concordance across all categories was 88%. Agreement analysis using weighted kappa statistics showed a linear weighted κ of 0.79 (95% confidence interval [CI], 0.75–0.84) and a quadratic weighted κ of 0.82 (95% CI, 0.78–0.86), indicating moderate-to-strong agreement between the two assays [22].

Discussion

Although Tg is a well-established biomarker for post-therapeutic surveillance of thyroid cancer, its utility in the initial diagnostic evaluation of thyroid nodules is limited. Routine measurement of serum Tg is not recommended during the initial workup because of its low specificity [23,24]. Elevated levels may also be observed in benign conditions such as multinodular goiter and hyperthyroidism. However, in cases with indeterminate cytology, elevated preoperative Tg levels may provide additional diagnostic insights. A recent meta-analysis demonstrated that patients with DTC exhibit significantly elevated Tg levels, with a 2.6-fold increase in the odds of malignancy compared to individuals with benign thyroid nodules [10]. The American Thyroid Association categorizes treatment responses for DTCs into four distinct categories. An excellent response is characterized by a suppressed Tg concentration (<0.2 ng/mL) or stimulated Tg level (<1 ng/mL) with no clinical, biochemical, or structural evidence of disease [25]. Although Tg has limited utility in the initial diagnosis of thyroid nodules due to its lack of specificity, it plays a critical role in the posttreatment surveillance of DTC.

As shown in Fig. 2, the Tg values measured using the Cobas Elecsys Tg II assay were consistently higher across most evaluated concentration ranges, with a mean positive bias of approximately 3.7 ng/mL compared to the Atellica IM Tg assay. Notably, as shown in Table 2, the assay concordance in the high Tg concentration group was substantially lower than that in the low and normal Tg groups. This indicates increased variability between the two assays at elevated Tg concentrations. Consistent with the recommendations of the American Thyroid Association and other expert panels, it is strongly advised that serial Tg measurements for individual patients be performed using the same assay platform to ensure reliable trend interpretation [26]. Notably, previous research has demonstrated that inter-assay variability may result in discordant clinical classifications in up to 7% of patients, emphasizing the potential risk of misinterpretation when transitioning between different Tg assay systems [27]. It is thus important to recognize that assay-specific cutoffs may influence clinical performance [28].

Although the overall correlation remained excellent, with a Pearson correlation coefficient of 0.997, the concordance between the two assays was notably lower within the high concentration range. This discrepancy may be attributable to differences in the predefined clinical cutoffs and reference intervals of each assay, which could influence the classification outcomes despite strong linear relationships. In a comparative study of two serum-free light-chain assays, Smith and Wu [29] reported a strong linear correlation between the methods, with Pearson’s correlation coefficients ranging from 0.95 to 0.97. Despite this high correlation, the overall concordance in the categorical classification was approximately 90% when the manufacturer-specific reference intervals of each assay were applied. Most discordant results occurred near the upper decision threshold. Adjusting the upper reference limit for free lambda light chains to 32.0 mg/L (from 27.69) increased concordance to 94%, suggesting that discrepancies were largely attributable to the cutoff differences rather than analytical imprecision [29]. The Atellica IM Tg assay demonstrated a CV within the manufacturers claims, with a repeatability of 1.4%–3.3% and within-lab variability of 2.5%–3.3% (Table 1). Another contributing factor to the observed discrepancies at higher Tg concentrations is the difference in the AMR reference interval thresholds between the two assays. The Cobas Elecsys Tg II assay offers a broader upper measurement limit, extending to 500 ng/mL, in contrast to the Atellica IM Tg assay, which has an upper limit of 150 ng/mL. Additionally, the Atellica assay has a higher upper reference limit for normal Tg values (111 ng/mL) than the Cobas assay (77 ng/mL). These differences in the analytical span and clinical cutoff definitions may influence the assay concordance, particularly at elevated Tg levels. Although this hypothesis was not directly evaluated in the present study, it may partially explain the reduced agreement observed at high concentrations.

Although this study provides important comparative data, it has some limitations. First, the analysis was confined to two Tg assay platforms, Cobas Elecsys Tg II and Atellica IM Tg, which may limit the extrapolation of the findings. Despite the strong correlation between the two methods, the absence of assay harmonization or the use of a standardized reference material may have contributed to the interassay discrepancies, particularly in the higher concentration range. The lack of evaluation of anti-Tg antibodies (TgAbs), which are known to interfere with immunometric Tg assays, could affect the interpretation of results. Studies with a lack of clinical follow-up data were excluded, limiting the ability of the assessment to observe differences that influence clinical decision making or patient outcomes. As no clinical information associated with patient samples was available, the present study was restricted to evaluating analytical rather than clinical performance. Finally, the small sample size within the high Tg concentration subgroup may have reduced the robustness of the concordance analysis at elevated Tg levels.

This study demonstrated a strong overall correlation between the Roche Cobas Elecsys Tg II and Siemens Atellica IM Tg assays for serum Tg measurements, supporting their analytical alignment across a wide range of concentrations. Nonetheless, significant interassay variability was evident, particularly at elevated Tg levels, resulting in reduced concordance and the potential for discordant clinical categorization. Given the central role of Tg in the long-term monitoring of patients with DTC, these findings underscore the importance of maintaining consistency in the assay used for serial measurements. The interpretation of Tg trends across different assay platforms should be approached with caution, as methodological variations may adversely affect clinical decisions. Further investigations incorporating TgAb status and long-term clinical outcomes are warranted to streamline harmonization strategies across testing platforms.

Article information

Conflicts of interest

Hyunyong Hwang is an editorial board member of the journal but was not involved in the peer reviewer selection, evaluation, or decision process of this article. The funder had no role in study design, analysis, interpretation or publication. We declare that we have no financial conflicts of interest.

Funding

This research was supported by Siemens Healthineers company in 2023.

Author contributions

Conceptualization: HH. Data curation; Formal analysis: all authors. Funding acquisition: HH. Investigation; Methodology: all authors. Project administration; Resources; Supervision: HH. Validation; Visualization: all authors. Writing-original draft: KHR. Writing-review & editing: HH. All authors read and approved the final manuscript.

Fig. 1.

Linearity assessment of the Siemens Atellica IM thyroglobulin assay. (A) Linearity data versus linearity fit. (B) Comparison of ADL to the deviation from linearity. ADL, allowable deviation from linearity; CI, confidence interval.

Fig. 2.

Comparison of Siemens Atellica IM Tg and Roche Cobas Elecsys Tg II assays using Bland-Altman analysis (blue line) and Passing-Bablok regression (red line). (A) Bland-Altman plot assessing agreement. (B) Passing-Bablok regression for correlation. Tg, thyroglobulin; LoA, limits of agreement.

References

• 1. Jung KW, Kang MJ, Park EH, Yun EH, Kim HJ, Kim JE, et al. Prediction of cancer incidence and mortality in Korea, 2024. Cancer Res Treat 2024;56:372–9.ArticlePubMedPMCPDF
• 2. Kang MJ, Jung KW, Bang SH, Choi SH, Park EH, Yun EH, et al. Cancer statistics in Korea: incidence, mortality, survival, and prevalence in 2020. Cancer Res Treat 2023;55:385–99.ArticlePubMedPMCPDF
• 3. Haugen BR, Alexander EK, Bible KC, Doherty GM, Mandel SJ, Nikiforov YE, et al. 2015 American Thyroid Association Management Guidelines for adult patients with thyroid nodules and differentiated thyroid cancer: the American Thyroid Association Guidelines Task Force on thyroid nodules and differentiated thyroid cancer. Thyroid 2016;26:1–133.ArticlePubMedPMC
• 4. Chen P, Yao Y, Tan H, Li J. Systemic treatments for radioiodine-refractory thyroid cancers. Front Endocrinol (Lausanne) 2024;15:1346476.ArticlePubMedPMC
• 5. Lee JC, Song BS, Kang YM, Kim YR, Kang YE, Lee JH, et al. Effect of thyroid-stimulating hormone suppression on muscle function after total thyroidectomy in patients with thyroid cancer. Front Endocrinol (Lausanne) 2021;12:769074.ArticlePubMedPMC
• 6. Indrasena BS. Use of thyroglobulin as a tumour marker. World J Biol Chem 2017;8:81–5.ArticlePubMedPMC
• 7. Prpic M, Franceschi M, Romic M, Jukic T, Kusic Z. Thyroglobulin as a tumor marker in differentiated thyroid cancer: clinical considerations. Acta Clin Croat 2018;57:518–27.ArticlePubMedPMC
• 8. Giovanella L, D'Aurizio F, Algeciras-Schimnich A, Gorges R, Petranovic Ovcaricek P, Tuttle RM, et al. Thyroglobulin and thyroglobulin antibody: an updated clinical and laboratory expert consensus. Eur J Endocrinol 2023;189:R11–27.ArticlePubMedPDF
• 9. Peiris AN, Medlock D, Gavin M. Thyroglobulin for monitoring for thyroid cancer recurrence. JAMA 2019;321:1228.ArticlePubMed
• 10. Lu Y, Zhao H, Liu C, Kuang Z, Li X. The role of preoperative serum thyroglobulin in the diagnosis and treatment of differentiated thyroid cancer: a systematic review and meta-analysis. Front Oncol 2024;14:1426785.ArticlePubMedPMC
• 11. Kraus T, Shengelia-de Lange N, Einspieler H, Hacker M, Haug A, Kretschmer-Chott E, et al. Value of follow-up diagnostic radioiodine scans in differentiated thyroid cancer. Endocr Connect 2024;13:e240007.ArticlePubMedPMC
• 12. Moosavi M, Kreisman S. A case report of dramatically increased thyroglobulin after lymph node biopsy in thyroid carcinoma after total thyroidectomy and radioiodine. Case Rep Endocrinol 2016;2016:6471081.ArticlePubMedPMCPDF
• 13. Highly sensitive thyroglobulin assay is useful in the follow-up of patients treated for DTC. Nat Clin Pract Endocrinol Metab 2007;3:790.ArticlePDF
• 14. Maus A, Thompson C, Grebe SK. Measurements of iodination in thyroglobulin: a step toward the next generation of thyroid cancer monitoring. J Endocr Soc 2025;9:bvaf015.ArticlePubMedPMCPDF
• 15. Clark P, Franklyn J. Can we interpret serum thyroglobulin results? Ann Clin Biochem 2012;49:313–22.ArticlePubMedPDF
• 16. Aguirre JJ, Ness K, Algeciras-Schimnich A. Application of the CLSI EP15-A3 guideline as an alternative troubleshooting tool for verification of assay precision. Am J Clin Pathol 2019;152(Suppl_1):S88.Article
• 17. Jeong S, Hong YR, Hwang H. Performance comparison between Elecsys anti-SARS-CoV-2 and anti-SARS-CoV-2 S and Atellica IM SARS-CoV-2 total and SARS-CoV-2 IgG assays. Kosin Med J 2022;37:154–62.ArticlePDF
• 18. Choi H, Lee SS, Hwang H. Comparative analysis of access PCT and Elecsys BRAHMS PCT assays for procalcitonin measurements. Kosin Med J 2024;39:272–80.ArticlePDF
• 19. Cho J. Evaluating and establishing the linearity interval and extended measuring interval. Lab Med Online 2024;14:163–75.Article
• 20. Garrett PE, Lasky FD, Meier KL, Clark LW; Clinical and Laboratory Standards Institute. User protocol for evaluation of qualitative test performance: approved guideline. 2nd ed. Clinical and Laboratory Standards Institute; 2008.
• 21. Ozarda Y, Higgins V, Adeli K. Verification of reference intervals in routine clinical laboratories: practical challenges and recommendations. Clin Chem Lab Med 2018;57:30–7.ArticlePubMed
• 22. McHugh ML. Interrater reliability: the kappa statistic. Biochem Med (Zagreb) 2012;22:276–82.ArticlePubMedPMC
• 23. Lee EK, Park YJ, Jung CK, Na DG. A narrative review of the 2023 Korean Thyroid Association Management Guideline for patients with thyroid nodules. Endocrinol Metab (Seoul) 2024;39:61–72.ArticlePubMedPMCPDF
• 24. Uludag M, Unlu MT, Kostek M, Aygun N, Caliskan O, Ozel A, et al. Management of thyroid nodules. Sisli Etfal Hastan Tip Bul 2023;57:287–304.ArticlePubMedPMC
• 25. Sipos JA, Aloi J, Gianoukakis A, Lee SL, Klopper JP, Kung JT, et al. Thyroglobulin cutoff values for detecting excellent response to therapy in patients with differentiated thyroid cancer. J Endocr Soc 2023;7:bvad102.ArticlePubMedPMCPDF
• 26. Giovanella L, Clark PM, Chiovato L, Duntas L, Elisei R, Feldt-Rasmussen U, et al. Thyroglobulin measurement using highly sensitive assays in patients with differentiated thyroid cancer: a clinical position paper. Eur J Endocrinol 2014;171:R33–46.ArticlePubMedPDF
• 27. Schoonen L, Neele M, van Toor H, van Kinschot CM, van Noord C, Visser WE, et al. Impact of thyroglobulin and thyroglobulin antibody assay performance on the differential classification of DTC patients. J Endocr Soc 2022;6:bvab166.ArticlePubMedPMCPDF
• 28. Katrangi W, Grebe SK, Algeciras-Schimnich A. Analytical and clinical performance of thyroglobulin autoantibody assays in thyroid cancer follow-up. Clin Chem Lab Med 2017;55:1987–94.ArticlePubMed
• 29. Smith A, Wu AH. Analytical and clinical concordance of free light chain assay. Pract Lab Med 2019;13:e00112.ArticlePubMedPMC

Figure & Data

References

Citations

Citations to this article as recorded by