Abstract
Context:
Measurement of thyroglobulin (Tg) by mass spectrometry (Tg-MS) is emerging as a tool for accurate Tg quantification in patients with anti-Tg autoantibodies (TgAbs).
Objective:
The objective of the study was to perform analytical and clinical evaluations of two Tg-MS assays in comparison with immunometric Tg assays (Tg-IAs) and Tg RIAs (Tg-RIAs) in a cohort of thyroid cancer patients.
Methods:
A total of 589 samples from 495 patients, 243 TgAb−/252 TgAb+, were tested by Beckman, Roche, Siemens-Immulite, and Thermo-Brahms Tg and TgAb assays, two Tg-RIAs, and two Tg-MS assays.
Results:
The frequency of TgAb+ was 58%, 41%, 27%, and 39% for Roche, Beckman, Siemens-Immulite, and Thermo-Brahms, respectively. In TgAb− samples, clinical sensitivities and specificities of 100% and 74%–100%, respectively, were observed across all assays. In TgAb+ samples, all Tg-IAs demonstrated assay-dependent Tg underestimation, ranging from 41% to 86%. In TgAb+ samples, the use of a common cutoff (0.5 ng/mL) for the Tg-MS, three Tg-IAs, and the USC-RIA improved the sensitivity for the Tg-MSs and Tg-RIAs when compared with the Tg-IAs. In up to 20% of TgAb+ cases, Tg-IAs failed to detect Tg that was detectable by Tg-MS. In Tg-RIAs false-high biases were observed in TgAb+ samples containing low Tg concentrations.
Conclusions:
Tg-IAs remain the method of choice for Tg quantitation in TgAb− patients. In TgAb+ patients with undetectable Tg by immunometric assay, the Tg-MS will detect Tg in up to 20% additional cases. The Tg-RIA will detect Tg in approximately 35% cases, but a significant proportion of these will be clinical false-positive results. The undetectable Tg-MS seen in approximately 40% of TgAb+ cases in patients with disease need further evaluation.
Antithyroglobulin autoantibodies (TgAbs) are present in 20%–30% of patients with differentiated thyroid carcinoma (1–3). TgAbs can interfere with thyroglobulin (Tg) measurements, potentially resulting in false low results in Tg immunometric assays (Tg-IA), and false-high, or sometimes low, results in Tg RIAs (Tg-RIAs) (1, 4, 5). The problem is compounded by the fact that interference is variable between patients and Tg assays and is only loosely correlated with TgAb concentrations (6–10). Consequently, it is difficult in TgAb-positive (TgAb+) samples to predict the degree of interference, let alone to provide an accurate Tg result. Some cases with interference might also be completely missed because any given TgAb assay will detect only a subset of interfering TgAbs (10–12).
Recently mass spectrometry-based Tg quantification (Tg-MS) has emerged as a potential solution to these problems. The published methods are based on peptide quantitation after tryptic digestion and immunocapture of Tg-specific peptide(s) (13–20). Trypsin cleaves all proteins, including TgAbs and other antibodies, thus eliminating them as interferences. As a result, Tg-MS should allow accurate Tg quantification in TgAb+ samples (11–13, 18). However, it remains to be determined to what degree Tg-MS improves actual patient care in TgAb+ samples in comparison with Tg-IAs or Tg-RIAs; the latter possibly detects some relevant Tg elevations that are Tg negative (Tg−) by Tg-MS (10).
To establish the potential impact of Tg-MS on patient care, we performed a comprehensive analytical and clinical evaluation of two Tg-MS, four Tg-IA, two Tg-RIA, and four TgAb assays in a cohort of 495 well-characterized thyroid cancer patients. This allowed us to assess the performance of different Tg assays in the presence of TgAbs, resulting in an improved understanding of the strengths and weaknesses of several widely used assays and allowing us to potentially improve the interpretation of Tg measurements in TgAb+ patients.
Materials and Methods
Patients, samples, and overall study design
This study was reviewed and approved by the Mayo Clinic Institutional Review Board.
The study consisted of the following: 1) the evaluation of the calibration accuracy of various Tg assays and assessment of their behavior when exposed to standardized mixtures of Tg and TgAbs (Supplemental Methods and Supplemental Table 1); and 2) the analytical and clinical comparison of Tg and TgAb assays using well-characterized patient samples.
For the second part of the study, we aimed to accrue serum specimens from 500 thyroid cancer patients; half TgAb negative (TgAb−) and half TgAb+ (Roche assay; Roche Diagnostics), with approximately equal numbers of Tg+ and Tg− patients (Beckman assay; Beckman Coulter). Tg+ and Tg− for the Tg-IA, Tg-MS, and Tg-RIA assays were defined as outlined in the Supplemental Material. For Tg+ and TgAb+ samples, we tried to cover a large concentration range (Supplemental Table 2).
Samples were chosen from approximately1000 residual serum samples from Mayo Clinic Rochester patients who had undergone clinical Tg and TgAb testing during the previous 2 years. Samples were included if the patient had a diagnosis of thyroid cancer, if the residual sample volume was greater than 0.5 mL and if the sample had undergone one or fewer freeze/thaw cycles. Samples fulfilling these criteria were further selected to obtain an approximately equal number of TgAb− and TgAb+ samples to span the analytical range of the assay as indicated above. All Tg and TgAb measurements except in 42 samples (7%) were unstimulated measurements (on T4).
The final set included 589 samples from 495 unique patients (Supplemental Table 2). The medical records of all patients were reviewed to extract a standardized set of data (Supplemental Table 3). A disease status was assigned by three clinical endocrinologists (G.C.L., M.R.C., and A.F.T.) for each visit that resulted in a Tg blood draw that was included in the study. Assignment of disease status was based on the longitudinal review of the clinical, imaging, and biochemical (including Tg and TSH) data and any cytology/histology data available. Patients were classified as alive, free of disease (AFD) if there was no clinical, imaging, or cytological/histological evidence of disease at the time of the blood draw and their serum Tg was either less than 0.1 ng/mL (TgAb−) or, if it was higher (<2 ng/mL; in patients with a small remnant), it had remained unchanged over several visits. Patients who did not fulfill these criteria were classified as alive with disease (AWD) and further stratified into alive with locoregional disease (AWLD) or alive with distant disease (AWDD) based on disease extent.
Assays
Automated immunoassays
The Beckman Access Tg and TgAb (Beckman Coulter), Siemens-Immulite Tg and TgAb (Siemens), Thermo-Brahms Tg and TgAb (Thermo Scientific), Roche Elecsys Tg II and TgAb (Roche Diagnostics) (Supplemental Table 4) were used at the Mayo Clinic per the manufacturers' instructions.
These Tg assays are standardized against the certified reference material (BCR 457; European Commission Institute for Reference Materials and Measurements, Geel, Belgium); the TgAb assays are standardized against the World Health Organization 65/93 International Standard.
Liquid chromatography, tandem mass spectrometry assays
Tg-MS-1 was performed at the Mayo Clinic (Rochester, Minnesota) and Tg-MS-2 at the University of Washington (Seattle, Washington). Both assays are standardized against BCR 457; for assay details see Supplemental Methods.
Radioimmunoassays
Tg-RIAs were performed at the University of Southern California (USC-RIA) (Los Angeles, California) and the Queen Elizabeth Hospital (UK-RIA) (Birmingham, United Kingdom). These RIAs are standardized against BCR 457; for assay details, see Supplemental Methods.
Testing procedure
All samples were deidentified, and the participating laboratories were blinded to the original clinical testing results and associated clinical data.
All samples were tested by Tg-MS-1 and the four Tg-IA and TgAb assays. Samples with sufficient residual material were then distributed to the other sites for testing by Tg-MS-2, USC-RIA, and UK-RIA.
Assessment of analytical and clinical assay performance
The effects of TgAb status on Tg results were compared between Tg-MS and the other assays. This included the comparison of samples with measurable Tg by both Tg-MS and the respective other assay(s) as well as the comparison of samples detectable by one method but not the other. For these analyses, a sample was considered TgAb+ if it had tested positive with at least one assay.
Clinical performance was assessed based on correct identification of disease status by the various assays. Clinical test sensitivity and specificity were calculated. To avoid confounding by Tg secreted by remnant normal thyroid tissue, only patients who had undergone total thyroidectomy and radioiodine remnant ablation were included in the evaluation of clinical sensitivity and specificity. Each blood draw was considered an independent event, and hence, two samples coming from one patient at different draw times were considered as two independent cases.
Further analysis included evaluation of reflexing TgAb+, Tg− samples from Tg-IA to Tg-MS or Tg-RIA and analysis of clinical status of patients with discordant Tg-IA and Tg-MS result.
Data analysis
Categorical data were analyzed by χ2 testing or z-tests for proportions, whereas continuous data were analyzed by parametric or nonparametric tests for differences in means or medians, respectively (Excel 2010; Microsoft; and JMP Pro 10; SAS Institute). The receiver-operator characteristic (ROC) curves were generated using Analyze-it (Analyze-it Ltd).
Results
Assessment of assay calibration and performance in standardized Tg/TgAb admixtures
In the absence of TgAbs, the r2 for linear calibration fit was greater than 0.99 for all comparisons, but there were slope biases between the assays (Supplemental Figure 1A). There was no significant effect of TgAb+ on Tg measurements by Tg-MS, marked negative biases in all Tg-IA with increasing TgAb concentrations, and marked positive biases in both Tg-RIAs at the highest TgAb/lowest Tg (1 ng/mL) concentrations and negative Tg-RIA biases at the highest TgAb/highest Tg (100 ng/mL) concentrations (Supplemental Figure 2).
Tg and TgAb concentrations in thyroid cancer patients
Most patients had stage I papillary thyroid carcinomas at initial diagnosis and were AFD at the time of blood draw (Table 1).
Table 1.
Patient Demographics and Clinical Characteristics
| Clinical Information | n |
|---|---|
| Unique patients (total samples) | 495 (589) |
| Sex | 346 F/149 M |
| Age median (range), y | 52 (13–89) |
| Disease status at time of drawa | |
| AFD | 377 |
| AWLDb | 103 |
| AWDDb | 86 |
| Recent diagnosis (<6 mo) | 23 |
| Morphotype | |
| PTC | 443 |
| FTC | 21 |
| HCC | 21 |
| Anaplastic | 10 |
| Stagingc | |
| I | 282 |
| II | 27 |
| III | 90 |
| IV | 81 |
| Unknown | 15 |
| Type of surgery | |
| Total/near thyroidectomy | 471 |
| Subtotal thyroidectomy | 10 |
| Hemithyroidectomy or less | 7 |
| No surgery | 7 |
| Radioactive iodine remnant ablation | |
| Yes | 215 |
| No | 278 |
| Unknown | 2 |
| Median follow-up, y (range) | 7.3 (0.05–46.7) |
Abbreviations: F, female; FTC, follicular thyroid carcinoma; HCC, Hurthle cell carcinoma; M, male; PTC, papillary thyroid carcinoma.
Patients who had a change in disease status over the course of sample collection were counted in more than one category.
For data analysis purposes, the AWLD and AWDD were combined.
Staging based on the AJCC Cancer Staging Manual, seventh edition (27).
Except for the Siemens-Immulite, which showed a significantly lower frequency of Tg+ (P < .0001), no differences were observed between the different Tg-IAs for the mean/median Tg values or for the frequencies of Tg+ samples (P ≥ .27) (Supplemental Table 5). The number of TgAb+ samples varied between assays: Roche, n = 339 (58%); Beckman, n = 241 (41%); Siemens-Immulite, n = 158 (27%); and Thermo-Brahms, n = 227 (39%) (Supplemental Table 6); P < .0001 for all assay comparisons, except Beckman to Thermo-Brahms assay (P > .9). The different TgAb assays did not consistently identify the same samples as TgAb+; TgAbs could not be detected by any TgAb assay in 229 samples (39%), detectable by only one of the four assays in 87 samples (15%), by two of the four assays in 77 samples (13%), by three of the four assays in 62 samples (10%), and by all four assays in 134 samples (23%).
Tg assay concordance in TgAb− and TgAb+ cases
The overall concordance between the Tg-IAs and Tg-MS-1 in TgAb+ samples is shown in Table 2. The concordance of negative results between Tg-IAs and Tg-MS-1 was 86%–94% when the functional sensitivity (FS) of each assay was used to determined positivity. This concordance improved to 95%–99% if a 0.5 ng/mL negative cutoff was used for the Tg-IAs, indicating that most cases classified as negative by Tg-MS-1 and positive by Tg-IAs had values between 0.1 and 0.5 ng/mL, ie, the discrepancies were due to the differences in the assays' FS.
Table 2.
Concordance Between Tg-MS, Tg-IA, and Tg-RIA in TgAb+ Samples
| Tg-IA or Tg-MS, ng/mL | Tg-MS-1, ng/mL |
|
|---|---|---|
| Negative <0.5 | Positive >0.5 | |
| Roche | ||
| Negative <0.1 | 197 | 16 |
| Positive >0.1 | 31 | 106 |
| Negative <0.5 | 223 | 27 |
| Positive >0.5 | 5 | 95 |
| Beckman | ||
| Negative <0.1 | 191 | 16 |
| Positive >0.1 | 44 | 109 |
| Negative <0.5 | 230 | 36 |
| Positive >0.5 | 5 | 89 |
| Siemens-Immulite | ||
| Negative <0.9 | 232 | 56 |
| Positive >0.9 | 3 | 68 |
| Thermo-Brahms | ||
| Negative <0.15 | 195 | 12 |
| Positive >0.15 | 39 | 112 |
| Negative <0.5 | 232 | 32 |
| Positive >0.5 | 2 | 92 |
| Tg-MS-2 | ||
| Negative <0.5 | 44 | 10 |
| Positive >0.5 | 0 | 17 |
| USC-RIA | ||
| Negative <0.5 | 16 | 2 |
| Positive >0.5 | 39 | 26 |
| UK-RIA | ||
| Negative <5.0 | 27 | 14 |
| Positive >5.0 | 27 | 15 |
Analysis was performed using the assays' FS and a 0.5 ng/mL cutoff for assays with FS less than 0.5 ng/mL.
Comparison of Tg measurements between Tg-MS-1 and Tg-IAs for the TgAb− and TgAb+ samples that contained 0.5 ng/mL or greater of Tg (FS of Tg-MS-1) by both Tg-IA and Tg-MS-1 showed that the Tg-IAs correlated well with Tg-MS-1 in TgAb−, albeit with variable positive biases (Figure 1 and Supplemental Figure 3). Slope biases were also observed when comparing the Tg-IAs with each other in TgAb− and TgAb+ samples (Supplemental Figure 4). The difference observed in the TgAb− patients was in agreement with the bias observed in the assay calibration assessment (Supplemental Figure 1A). By contrast, the TgAb+ samples displayed negative biases against their TgAb− counterparts and against Tg-MS-1 (Figure 1). The median TgAb-induced negative biases relative to TgAb− samples were as follows: 41% (Beckman), 42% (Thermo-Brahms), 50% (Roche), and 86% (Siemens-Immulite); however, the spread of the bias varied between the Tg-IAs (Figure 1). On a cross-sectional analysis, the degree of Tg underestimation did not correlate with TgAb concentrations (r2 = 0.037) (Supplemental Figure 5A). Longitudinal analysis of individual patients with three or more samples gave a mixed picture. A relationship between Tg underestimation and TgAb concentration was observed in only some patients (Supplemental Figure 5B).
Figure 1.
Tg-IA bias in TgAb-positive samples. Folded probability (Mountain) plots compare the biases of the Tg-IAs to Tg-MS-1 in TgAb− (black) and TgAb+ (blue) samples with Tg concentrations greater than 0.5 ng/mL by Tg-MS-1 and by the respective Tg-IAs. The number of samples analyzed in each group were as follows: 109 TgAb− and 106 TgAb+ for Roche (panel A); 110 TgAb− and 109 TgAb+ for Beckman (panel B); 108 TgAb− and 93 TgAb+ for Siemens-Immulite (panel C); and 110 TgAb− and 112 TgAb+ for Thermo-Brahms (panel D). The x-axis of each plot shows the percentage difference between Tg-MS-1 and the respective Tg-IA assay. The y-axis depicts the probability of the occurrence of the observed differences. The peak of the mountain is at 50% probability, ie, the median percentage difference to Tg-MS-1. The left slope corresponds to the distribution of less positive, or greater negative, differences, whereas the right slope shows those of less negative, or greater positive, differences, respectively. In each panel, the light gray line denotes a (theoretical) 0% bias to Tg-MS-1; if there was complete concordance between Tg-MS-1 and a Tg-IA, the mountain would be a single line superimposed on this gray line. The black and blue dashed lines correspond to the actual Tg-MS-1 vs Tg-IA median biases in TgAb− and TgAb+ samples, respectively. The difference in percentage bias between TgAb− and TgAb+ samples are indicated by the arrows between the two peaks, and their numerical values are listed.
Concordance between the Tg-MS assays was evaluated in a subset of the samples (40 TgAb− and 71 TgAb+). Comparison between Tg-MS assays in TgAb− samples with a Tg concentration of 0.5 ng/mL or greater showed excellent agreement (Supplemental Figure 1B). In the TgAb− cases, there was 100% concordance between the assays for Tg+ results and 91% agreement for Tg− results (Table 2); two cases were positive by Tg-MS-1 (Tg = 0.65 and 0.79 ng/mL) but negative by Tg-MS-2. In the TgAb+ cases, the corresponding concordance percentages were 100% and 81%, respectively, with 10 cases (14%) positive by Tg-MS-1 (Tg range 0.54–2.5 ng/mL) but negative by Tg-MS-2.
Performance of Tg-MS reflex testing
If the Tg-MS-1 assay had been used as a reflex strategy for Tg-/TgAb+ samples, Tg would have been detected in an additional 51 Tg-Ab+ patients that were Tg− by one or more Tg-IA methods (Table 3, lines 1–51). Tg-MS-1 detected Tg in an additional 6% of samples compared with the Thermo-Brahms Tg-IA (12 of 207), 8% for the Beckman Tg-IA (16 of 207), 8% for the Roche Tg-IA (16 of 213), and 20% for the Siemens-Immulite Tg-IA (56 of 288) (Table 2). It should be noted that there were five samples that were TgAb− by all methods with detectable Tg by Tg-MS-1 but undetectable by at least one Tg-IA. Three were undetectable by all Tg-IAs and two were undetectable by the Siemens-Immulite Tg-IA (Table 3, lines 52–56).
Table 3.
TgAb+ Patients With Tg-MS Greater Than FS and Immunoassay Less Than FS
| ID | Disease Status | Tg, ng/mLa |
Anti-TgAb Assay, IU/mL |
|||||||
|---|---|---|---|---|---|---|---|---|---|---|
| LC-MS | Roche | Beckman | Immulite | Brahms | Roche | Beckman | Immulite | Brahms | ||
| 1 | AFD | 0.80 | <0.1 | 0.1 | <0.9 | <0.15 | 1565 (+) | 75 (+) | 735 (+) | 1167 (+) |
| 2 | AFD | 1.00 | 0.12 | 0.1 | <0.9 | <0.15 | 38 (+) | <1.8 (−) | <20 (−) | <30 (−) |
| 3 | AFD | 0.54 | 0.89 | 0.8 | <0.9 | 0.71 | 273 (+) | 6.2 (+) | 20 (+) | 149 (+) |
| 4 | AFD | 1.48 | <0.1 | 1.3 | <0.9 | <0.15 | 37 (+) | <1.8 (−) | <20 (−) | <30 (−) |
| 5 | AFD | 0.66 | 0.16 | 0.1 | <0.9 | 0.3 | 170 (+) | 345 (+) | 1128 (+) | <30 (−) |
| 6 | AFD | 0.50 | <0.1 | <0.1 | <0.9 | <0.15 | 51 (+) | <1.8 (−) | <20 (−) | <30 (−) |
| 7 | AFD | 0.82 | <0.1 | 0.1 | <0.9 | 1.85 | 91 (+) | 71 (+) | 271 (+) | 33 (+) |
| 8 | AFD | 0.59 | <0.1 | <0.1 | <0.9 | <0.15 | 23 (+) | <1.8 (−) | <20 (−) | 34 (+) |
| 9 | AFD | 1.24 | 0.52 | 0.4 | <0.9 | 0.57 | 155 (+) | 3.4 (+) | <20 (−) | 40 (+) |
| 10 | AFD | 0.82 | <0.1 | 0.3 | <0.9 | 0.29 | 46 (+) | 2.9 (+) | <20 (−) | 42 (+) |
| 11 | AFD | 2.05 | 1.11 | 0.7 | <0.9 | 0.61 | 80 (+) | 24 (+) | <20 (−) | 43 (+) |
| 12 | AFD | 0.66 | <0.1 | 0.1 | <0.9 | 0.15 | 92 (+) | 7.7 (+) | <20 (−) | 52 (+) |
| 13 | AFD | 2.82 | <0.1 | <0.1 | <0.9 | <0.15 | 554 (+) | 28 (+) | 133 (+) | 548 (+) |
| 14 | AFD | 1.80 | <0.1 | <0.1 | <0.9 | <0.15 | 42 (+) | 20 (+) | <20 (−) | 57 (+) |
| 15 | AFD | 0.82 | <0.1 | <0.1 | <0.9 | <0.15 | 36 (+) | 12 (+) | 45 (+) | <30 (−) |
| 16 | AFD | 0.71 | <0.1 | <0.1 | <0.9 | 0.19 | 29 (+) | 35 (+) | <20 (-) | <30 (−) |
| 17 | AWDD | 1.59 | 0.98 | 0.4 | <0.9 | 0.40 | >3000 (+) | 684 (+) | >3000 (+) | 10 590 (+) |
| 18 | AWDD | 0.54 | 0.20 | 0.3 | <0.9 | 0.20 | 26 (+) | <1.8 (−) | <20 (−) | <30 (−) |
| 19 | AWDD | 2.60 | 2.56 | <0.1 | 1.16 | 1.50 | 286 (+) | 412 (+) | 361 (+) | 1234 (+) |
| 20 | AWDD | 0.63 | 0.24 | 0.2 | <0.9 | 0.15 | 1047 (+) | 247 (+) | 663 (+) | 1378 (+) |
| 21 | AWDD | 0.70 | <0.1 | <0.1 | <0.9 | <0.15 | 285 (+) | 7.5 (+) | 36 (+) | 157 (+) |
| 22 | AWDD | 1.37 | 1.08 | 0.9 | <0.9 | 0.90 | 456 (+) | 10 (+) | 42.9 (+) | 214 (+) |
| 23 | AWDD | 2.21 | 0.69 | 0.6 | <0.9 | 0.41 | >3000 (+) | 2262 (+) | 2533 (+) | 2789 (+) |
| 24 | AWDD | 0.60 | 0.16 | 1.1 | <0.9 | 0.17 | 66 (+) | <1.8 (−) | <20 (−) | <30 (−) |
| 25 | AWDD | 1.19 | NA | 0.6 | <0.9 | 0.76 | 911 (+) | 515 (+) | 546 (+) | 4444 (+) |
| 26 | AWDD | 1.48 | 1.26 | 0.9 | <0.9 | 0.95 | 89 (+) | <1.8 (−) | <20 (−) | 48 (+) |
| 27 | AWDD | 0.83 | 0.43 | 0.2 | <0.9 | 0.29 | 312 (+) | 3.8 (+) | <20 (−) | 50 (+) |
| 28 | AWDD | 16.60 | 2.50 | 2.5 | <0.9 | 2.12 | 541 (+) | 23 (+) | 88 (+) | 661 (+) |
| 29 | AWDD | 1.73 | 0.66 | 0.5 | <0.9 | 0.34 | >3000 (+) | 2262 (+) | >3000 (+) | NA |
| 30 | AWLD | 2.08 | 1.93 | 1.6 | <0.9 | 1.46 | 284 (+) | 4.2 (+) | <20 (−) | 102 (+) |
| 31 | AWLD | 0.50 | 0.29 | 0.3 | <0.9 | 0.20 | 25 (+) | 2.1 (+) | <20 (−) | <30 (−) |
| 32 | AWLD | 8.71 | <0.1 | <0.1 | <0.9 | 0.46 | 51 (+) | <1.8 (−) | <20 (−) | <30 (−) |
| 33 | AWLD | 0.76 | 0.72 | 0.7 | <0.9 | 0.64 | 532 (+) | 154 (+) | 100 (+) | 186 (+) |
| 34 | AWLD | 0.88 | <0.1 | 0.1 | <0.9 | <0.15 | 551 (+) | 144 (+) | 107 (+) | 195 (+) |
| 35 | AWLD | 0.67 | 0.18 | 0.1 | <0.9 | 0.43 | 28 (+) | 2.2 (+) | <20 (−) | <30 (−) |
| 36 | AWLD | 1.18 | 1.00 | <0.1 | <0.9 | 0.62 | 248 (+) | 279 (+) | 255 (+) | 215 (+) |
| 37 | AWLD | 1.74 | 0.99 | 0.9 | <0.9 | 0.68 | 37 (+) | <1.8 (−) | <20 (−) | <30 (−) |
| 38 | AWLD | 1.63 | 1.49 | 0.8 | <0.9 | 1.05 | 27 (+) | <1.8 (−) | <20 (−) | <30 (−) |
| 39 | AWLD | 1.05 | 0.71 | 0.4 | <0.9 | 0.50 | 22 (+) | <1.8 (−) | <20 (−) | <30 (−) |
| 40 | AWLD | 0.79 | 0.64 | 0.7 | <0.9 | 0.63 | 24 (+) | <1.8 (−) | <20 (−) | 364 (+) |
| 41 | AWLD | 0.80 | 0.63 | 0.5 | <0.9 | 0.51 | <20 (−) | <1.8 (−) | <20 (−) | 38 (+) |
| 42 | AWLD | 5.36 | 1.47 | 1.2 | <0.9 | 0.89 | 141 (+) | 5.5 (+) | 65 (+) | 46 (+) |
| 43 | AWLD | 0.99 | 0.59 | 0.7 | <0.9 | 0.54 | 28 (+) | <1.8 (−) | <20 (−) | <30 (−) |
| 44 | AWLD | 0.56 | 0.30 | 0.3 | <0.9 | 0.31 | 209 (+) | <1.8 (−) | <20 (−) | 59 (+) |
| 45 | AWLD | 0.81 | 0.40 | 0.4 | <0.9 | 0.30 | 872 (+) | 2081 (+) | 2498 (+) | 7153 (+) |
| 46 | AWLD | 1.76 | 1.13 | 0.9 | <0.9 | 0.78 | 838 (+) | 50 (+) | 219 (+) | 825 (+) |
| 47 | AWLD | 1.03 | 0.69 | 0.5 | <0.9 | 0.48 | 146 (+) | 2.2 (+) | <20 (−) | 92 (+) |
| 48 | RDx | 1.12 | 1.54 | 0.8 | <0.9 | 0.67 | 33 (+) | <1.8 (−) | <20 (−) | <30 (−) |
| 49 | RDx | 0.59 | <0.1 | <0.1 | <0.9 | <0.15 | 31 (+) | <1.8 (−) | <20 (−) | <30 (−) |
| 50 | RDx | 1.75 | 0.66 | 0.4 | <0.9 | 0.51 | 449 (+) | 24 (+) | 27 (+) | 185 (+) |
| 51 | RDx | 0.82 | 0.45 | 0.3 | <0.9 | 0.48 | 47 (+) | 17 (+) | 80 (+) | 32 (+) |
| 52b | AFD | 0.66 | <0.1 | <0.1 | <0.9 | <0.15 | <20 (−) | <1.8 (−) | <20 (−) | <30 (−) |
| 53b | AFD | 2.30 | <0.1 | <0.1 | <0.9 | <0.15 | <20 (−) | <1.8 (−) | <20 (−) | <30 (−) |
| 54b | AFD | 0.98 | 0.95 | 0.90 | <0.9 | 0.60 | <20 (−) | <1.8 (−) | <20 (−) | <30 (−) |
| 55b | AFD | 1.35 | <0.1 | <0.1 | <0.9 | <0.15 | <20 (−) | <1.8 (−) | <20 (−) | <30 (−) |
| 56b | AWDD | 0.78 | 0.41 | 0.40 | <0.9 | 0.54 | <20 (−) | <1.8 (−) | <20 (−) | <30 (−) |
Abbreviation: ID, identification; LC-MS, liquid chromatography and mass spectrometry; NA, not available; RDx, recent diagnosis (<6 mo).
Tg measurements were made without TSH stimulation.
Samples 52–56 were TgAb− by all assays but show discordant results between the Tg-MS and TgIAs.
We also evaluated each manufacturer's combination of Tg/TgAb assay because they are typically used together. Tg was detectable by Tg-MS-1 but was undetectable by the Siemens-Immulite Tg-IA in 56 samples; 30 (54%) Siemens-Immulite TgAb− and 26 (46%) TgAb+. The Roche Tg-IA had 19 samples with undetectable Tg, detectable by Tg-MS-1; three (16%) Roche TgAb− and 16 (84%) TgAb+. The Beckman Tg-IA had 19 samples with undetectable Tg, detectable by Tg-MS-1; seven (37%) Beckman TgAb− and 12 (63%) TgAb+. The Thermo-Brahms Tg assay had 15 undetectable Tg samples, detectable by Tg-MS-1; eight (53%) were Thermo-Brahms TgAb− and seven (47%) TgAb+.
Clinical performance of Tg assays
Of the 589 samples, 242 samples were from patients who had undergone total thyroidectomy and radioactive remnant ablation, with 157 classified as AFD and 85 as AWD at the time of Tg testing. This subset was included in the analysis of clinical sensitivity and specificity. The ROC curves for TgAb− and TgAb+ cases are shown for each assay in Figure 2. There were no significant differences between the various assays in the areas under the curve for TgAb− and TgAb+ samples.
Figure 2.
ROC curves and clinical sensitivity and specificity analysis for Tg-MS, Tg-IA, and Tg-RIA. TgAb− samples (A) and TgAb+ samples (B) are shown. There are no significant differences between the area under the curves (AUC) of the different assays. However, the smaller number of data points makes AUC assessments of the Tg-RIA ROC curves less reliable than those of the other assays. Sensitivity and specificity were calculated based on the Tg assays' individual FS and a common 0.5 ng/mL cutoff. *, The assay's FS were as follows: 0.1 ng/mL Beckman and Roche; 0.15 ng/mL Thermo-Brahms; 0.9 ng/mL Immulite; 0.5 ng/mL Tg-MS-1, Tg-MS-2, and USC-RIA; and 5.0 ng/mL UK-RIA. ^, Assays with an FS of 0.5 ng/mL or greater are omitted from this portion of the table because the change to a 0.5 ng/mL cutoff does not affect them.
In the TgAb− samples, there were no significant differences in the Tg assays' sensitivities and specificities (all P > .91). In the Tg-MS assays, Tg was detectable in 100% of AWD cases and in 16% (Tg-MS-1) and 6% (Tg-MS-2) of AFD cases (84% and 94% specificity, respectively) (Figure 2). In the Tg-IA, Tg was detectable in 97%–100% of AWD cases and in 21%–26% of AFD cases (74%–85% specificities). Specificity improved without compromising the sensitivity for three of the four Tg-IAs if a cutoff of 0.5 ng/mL was used instead of the assay's FS. For the RIAs, Tg was positive in 100% of AWD cases with specificities of 83%–100%.
In TgAb+ samples, there were no significant differences in the sensitivities for the Tg-MS assays, Roche, Beckman, and Thermo-Brahms Tg-IA and USC-RIA (all P > .85) when the respective assays' FS was used (Figure 2). The Siemens-Immulite Tg-IA showed significantly lower sensitivity compared with the other Tg-IAs, Tg-MS, and Tg-RIAs (all P < .007). When a cutoff of 0.5 ng/mL was used, the sensitivities of the Tg-MS assays were slightly superior to the Tg-IAs with comparable specificity, in agreement with the increased number of TgAb+ samples that had detectable Tg by Tg-MS. The clinical specificity of the Tg-RIAs was significantly lower (P < .016) than that of the Tg-MS assays. The Tg-MS assays were positive in 56% of AWD cases and in 15% (Tg-MS-1) and 6% (Tg-MS-2) of AFD cases. For the RIAs, Tg was positive in 60%–81% of AWD cases and in 42%–55% of AFD cases. The difference in the frequency of AFD cases that were Tg+ by the USC-RIA vs the Tg-MS-1 was statistically significant (P = .002). To investigate the differences observed in the specificity between the Tg-MS and Tg-RIAs in the TgAb+ cases, a subset of these samples was tested by the Tg-MS-2 assay to determine whether these differences were method dependent. In all cases, Tg was undetectable (<0.5 ng/mL) by both Tg-MS assays but detectable by the USC-RIA (Table 4).
Table 4.
TgAb+ Patients With Tg-MS Less Than FS and Tg-RIA Greater Than FS
| ID | Disease Status | Tg Assay, ng/mLa |
Anti-TgAb Assay, IU/mL |
||||||
|---|---|---|---|---|---|---|---|---|---|
| LC-MS-1 | LC-MS-2 | USC-RIA | UK-RIA | Roche | Beckman | Immulite | Brahms | ||
| 1 | AFD | <0.5 | <0.5 | 0.7 | <5.0 | 21 (+) | <1.8 (−) | <20 (−) | <30 (−) |
| 2 | AFD | <0.5 | 0.7 | <20 (−) | <1.8 (−) | <20 (−) | <30 (−) | ||
| 3 | AFD | <0.5 | 0.8 | 25 (+) | 17 (+) | <20 (−) | <30 (−) | ||
| 4 | AFD | <0.5 | 0.8 | <5.0 | 32 (+) | <1.8 (−) | <20 (−) | <30 (−) | |
| 5 | AFD | <0.5 | <0.5 | 0.9 | 127 (+) | <1.8 (−) | <20 (−) | 67 (+) | |
| 6 | AFD | <0.5 | 0.9 | 22 (+) | 40 (+) | <20 (−) | <30 (−) | ||
| 7 | AFD | <0.5 | 0.9 | <5.0 | 308 (+) | 3.1 (+) | <20 (−) | 103 (+) | |
| 8 | AFD | <0.5 | <0.5 | 1.1 | 11.5 | 56 (+) | <1.8 (−) | <20 (−) | 43 (+) |
| 9 | AFD | <0.5 | 1.1 | 45 (+) | <1.8 (−) | 67 (+) | <30 (−) | ||
| 10 | AFD | <0.5 | 1.1 | 367 (+) | 36 (+) | 43 (+) | 192 (+) | ||
| 11 | AFD | <0.5 | 1.2 | 236 (+) | 47 (+) | 28 (+) | 129 (+) | ||
| 12 | AFD | <0.5 | 1.3 | 33 (+) | <1.8 (−) | <20 (−) | 38 (+) | ||
| 13 | AFD | <0.5 | 1.3 | 506 (+) | 25 (+) | 130 (+) | 446 (+) | ||
| 14 | AFD | <0.5 | 1.3 | 947 (+) | 67 (+) | 1090 (+) | 1850 (+) | ||
| 15 | AFD | <0.5 | 1.4 | <20 (−) | <1.8 (−) | <20 (−) | <30 (−) | ||
| 16 | AFD | <0.5 | 1.4 | 335 (+) | 8.5 (+) | 20 (+) | 205 (+) | ||
| 17 | AFD | <0.5 | <0.5 | 1.7 | 29 (+) | 4.1 (+) | <20 (−) | <30 (−) | |
| 18 | AFD | <0.5 | <0.5 | 1.7 | 13 | 79 (+) | 36 (+) | 96 (+) | 44 (+) |
| 19 | AFD | <0.5 | <0.5 | 2.3 | 10.9 | 34 (+) | 2.9 (+) | <20 (−) | <30 (−) |
| 20 | AFD | <0.5 | <0.5 | 2.4 | 79 (+) | 39 (+) | 60 (+) | 67 (+) | |
| 21 | AFD | <0.5 | 2.6 | 35 (+) | 4.4 (+) | <20 (−) | 30 (+) | ||
| 22 | AFD | <0.5 | 2.7 | 212 (+) | 123 (+) | 151 (+) | <30 (−) | ||
| 23 | AFD | <0.5 | 3.2 | 49 (+) | 2.7 (+) | 21 (+) | 45 (+) | ||
| 24 | AFD | <0.5 | 3.7 | 96 (+) | 3.8 (+) | 27 (+) | 73 (+) | ||
| 25 | AFD | <0.5 | 4.7 | 151 (+) | 4.6 (+) | 89 (+) | <30 (−) | ||
| 26 | AFD | <0.5 | 4.7 | 324 (+) | 56 (+) | 48 (+) | 191 (+) | ||
| 27 | AFD | <0.5 | 4.8 | 657 (+) | 22 (+) | 137 (+) | 172 (+) | ||
| 28 | AFD | <0.5 | <0.5 | 23 | 91 (+) | 4.4 (+) | <20 (-) | 58 (+) | |
| 29 | AFD | <0.5 | <0.5 | 5.7 | 137 (+) | 14 (+) | <20 (−) | 88 (+) | |
| 30 | AFD | <0.5 | <0.5 | 12.7 | 574 (+) | 397 (+) | 144 (+) | 486 (+) | |
| 31 | AFD | <0.5 | 10.2 | 32 (+) | 5.1 (+) | <20 (−) | <30 (−) | ||
| 32 | AFD | <0.5 | 10 | 95 (+) | 1.2 (−) | 27 (+) | 74 (+) | ||
| 33 | AFD | <0.5 | 201 | 110 (+) | 16 (+) | 37 (+) | 56 (+) | ||
| 34 | AFD | <0.5 | 11.6 | 165 (+) | 28 (+) | 29 (+) | 60 (+) | ||
| 35 | AFD | <0.5 | 9.3 | 201 (+) | 12 (+) | 26 (+) | 93 (+) | ||
| 36 | AFD | <0.5 | 13.3 | 283 (+) | 4.8 (+) | 54 (+) | 126 (+) | ||
| 37 | AFD | <0.5 | 15.9 | 784 (+) | 50 (+) | 113 (+) | 542 (+) | ||
| 38 | AFD | <0.5 | 41 | 3001 (+) | 245 (+) | 405 (+) | 407 (+) | ||
| 39 | AFD | <0.5 | 58.6 | 3001 (+) | 375 (+) | 722 (+) | 853 (+) | ||
| 40 | AWDD | <0.5 | 1.3 | 7.7 | 769 (+) | 47 (+) | 232 (+) | 4705 (+) | |
| 41 | AWDD | <0.5 | 8.7 | 1691 (+) | 446 (+) | 1100 (+) | 1971 (+) | ||
| 42 | AWDD | <0.5 | 9.7 | 266 (+) | 2.7 (+) | 45 (+) | 62 (+) | ||
| 43 | AWLD | <0.5 | <0.5 | 0.7 | 8.7 | 420 (+) | 13 (+) | 62 (+) | 178 (+) |
| 44 | AWLD | <0.5 | 0.8 | <5.0 | 121 (+) | 7.3 (+) | <20 (−) | 63 (+) | |
| 45 | AWLD | <0.5 | 1 | 231 (+) | 5.1 (+) | <20 (−) | 180 (+) | ||
| 46 | AWLD | <0.5 | 1.4 | 471 (+) | 5.0 (+) | 28 (+) | 208 (+) | ||
| 47 | AWLD | <0.5 | 1.5 | 28 (+) | 7.0 (+) | <20 (−) | <30 (−) | ||
| 48 | AWLD | <0.5 | 1.5 | <5.0 | 30 (+) | 3.4 (+) | <20 (−) | <30 (−) | |
| 49 | AWLD | <0.5 | 2.2 | 190 (+) | 17 (+) | 31 (+) | 189 (+) | ||
| 50 | AWLD | <0.5 | 2.7 | 957 (+) | 718 (+) | 846 (+) | 2842 (+) | ||
| 51 | AWLD | <0.5 | <0.5 | 4 | 25.7 | 310 (+) | 21 (+) | 80 (+) | 114 (+) |
| 52 | AWLD | <0.5 | <0.5 | 4.1 | 28.5 | 290 (+) | 24 (+) | 54 (+) | 104 (+) |
| 53 | AWLD | <0.5 | <0.5 | 4.8 | 34.3 | 127 (+) | 18 (+) | 72 (+) | 92 (+) |
| 54 | AWLD | <0.5 | 6 | 1006 (+) | 198 (+) | 342 (+) | 993 (+) | ||
| 55 | AWLD | <0.5 | 6.9 | 28 (+) | 25 (+) | 22 (+) | <30 (−) | ||
| 56 | AWLD | <0.5 | 5.3 | 32 (+) | 5.9 (+) | <20 (−) | <30 (−) | ||
| 57 | AWLD | <0.5 | 25.1 | 450 (+) | 95 (+) | 81 (+) | 156 (+) | ||
| 58 | AWLD | <0.5 | 41.5 | 1000 (+) | 134 (+) | 218 (+) | 167 (+) | ||
| 59 | AWLD | <0.5 | 37.5 | 1375 (+) | 2262 (+) | 3001 (+) | 16 616 (+) | ||
Abbreviations: ID, identification; LC-MS, liquid chromatography and mass spectrometry.
Tg measurements were made without TSH stimulation.
Clinical status of patients with discordant Tg-MS and Tg-IA results
Sixteen of the 51 unique TgAb+ patients (31%) who were Tg+ by Tg-MS-1 and Tg− by at least one Tg-IA were classified as AFD at the time of Tg testing. Their Tg-MS-1 concentrations ranged from 0.5 to 2.82 ng/mL (Table 3). There were 31 of 51 patients AWD (69%) with a Tg-MS-1 of 0.5–16.6 ng/mL who were Tg-IA negative by one assay (n = 31), three assays (n = 3), or four assays (n = 1). Four cases were patients with a recent diagnosis (<6 mo) with Tg concentrations ranging from 0.59 to 1.75 ng/mL. In the five TgAb− patients with detectable Tg by Tg-MS-1, which was missed by at least one Tg-IA, four were AFD and one was AWD. The Tg concentrations ranged from 0.6 to 2.3 ng/mL.
Discussion
This study provides several key analytical and clinical findings which are relevant to clinical practice. First, every TgAb assay misses a significant proportion of interfering TgAb (16%–54%). Even when four TgAb assays are used, at least 1.5% of interfering TgAbs are not detected if Tg by mass spectrometry is used as the reference (Table 2). If only a single TgAb assay is used, this percentage is severalfold higher.
Second, the presence of TgAbs causes a false-low bias in all Tg-IAs tested (Figure 1 and Supplemental Figure 2). We have extended the previous observations (1, 4, 21) by defining the expected underestimation for each Tg-IA when compared with Tg-MS.
Third, the degree of underestimation, ranging from 41% to 86%, was Tg assay specific, rather than TgAb concentration dependent (Figure 1 and Supplemental Figure 2). This underestimation is substantial enough to result in undetectable Tg in 6%–20% of TgAb+ samples. The underestimation and the rate of undetectable Tg concentrations are about 2-fold lower for the Beckman, Thermo-Brahms, and Roche Tg-IAs compared with the Siemens-Immulite Tg-IA. The reason for these differences might be due to the latter's inferior functional sensitivity (0.9 ng/mL) or the use of a suboptimal number of capture or detection antibodies compared with the other assays, which increases the likelihood that TgAb will hinder the binding of Tg to capture or detection antibodies. By increasing the number of assay antibodies, the fraction of patient TgAb that can interfere is reduced, in particular if the assay antibodies are selected based on known epitopes favored by TgAb found in cancer vs thyroid autoimmunity patients (22).
These findings support the need for using an assay with optimal FS (10, 23) and the least susceptibility to TgAb interference. Unfortunately, based on the review of the 2014 Proficiency Testing/External Quality Assurance program results from the College of America Pathologists, this is not the current state of thyroglobulin testing. Of 290 participants, which includes United States and non-US laboratories, 56% (n = 163) use the Siemens-Immulite Tg-IA, 30% (n = 86) use the Beckman Tg-IA, and 14% (n = 41) use the Roche Tg-IA. No participants reported the use of the Thermo-Brahms Tg-IA.
Fourth, different Tg-MS assays appear to agree with each other (Supplemental Figure 1B and Table 2). Individually, several Tg-MS assays have already proven capable of achieving accuracy and reproducibility comparable with Tg-IAs and Tg-RIAs while being able to detect Tg in TgAb-positive samples (10, 13–15, 18). Showing once again on a common set of samples that there is good agreement between two different Tg-MS assays confirms their reliability for clinical practice.
Fifth, 3 of the 4 Tg-IAs overall detect more Tg-positive samples than Tg-MS. This advantage is due to their superior detection sensitivity. As expected, this comes with a sacrifice in specificity (Figure 2). However, the concordance analysis on TgAb+ samples showed that the Tg-MS assays detect Tg in an additional 6%–20% of TgAb+ samples compared with the immunometric assays, which is supported by the superior clinical sensitivity in the Tg-MS assays when a comparable specificity is used (Figure 2).
Sixth, clinical sensitivities and specificities in TgAb− samples are high for all assays. Both are substantially lower in TgAb+ patients, regardless of the assay used (Figure 2), suggesting that some patients might either have detectable Tg from normal remnant tissue or might have active disease that secretes little, or no, Tg that is recognized by any assay.
Finally, Tg-MS failed to detect Tg in approximately 40% of TgAb+ AWD cases as previously reported (10). Whether undetectable Tg-MS associated with detectable Tg-RIA in TgAb+ AWD cases represents the failure of Tg-MS assays to recognize unusual Tg variants in cancer patients, increased clearance of Tg-TgAb complexes, or the Tg-RIA giving the correct clinical answer (AWD) for the wrong reason (false positive in Tg−, TgAb+ samples), remains to be determined. Although the USC-RIA in particular detected some clinically relevant positives in this population, the reliability of these detectable Tg-RIA values should be questioned when TgAb is very high because false-high Tg-RIA results were found for admixtures containing high TgAb concentrations and no or low Tg (Supplemental Figure 2). This is consistent with other studies showing TgAb interference with Tg-RIA (5, 21). Clinically, the increased rate detection of patients with AWD by the USC-RIA comes at the price of an increase in clinical false-positive results.
Based on these findings, Tg-IAs with optimal FS should remain the frontline assay in TgAb− patients. In TgAb+ patients with detectable Tg by immunometric assay, Tg-MS might be used if accurate quantitation is deemed clinically necessary. It might also be worthwhile to determine whether the predictable bias observed in TgAb+ samples is reproducible in a different patient cohort, given the known TgAb heterogeneity; if this bias is reproducible, it might assist in estimating the true Tg-IA Tg result in a TgAb+ sample. In TgAb+ patients with undetectable Tg by Tg-IA, the use of Tg-MS will detect and accurately quantify additional Tg+ cases (6%–20%); however, the suboptimal functional sensitivity of the current generation of Tg-MS assays needs to be considered because patients with Tg between 0.1 and 0.5 ng/mL might yield false-negative results by current Tg-MS assays.
Many of the Tg-MS+, Tg-IA− patients have low Tg concentrations and are clinically free of detectable disease. Tg-RIA detects about 35% of additional Tg-positive cases but at the cost of about a doubling of clinical false-positive results. Tg-RIA results in these patients will also be highly variable between different Tg-RIAs and are unlikely to reflect the true Tg concentration, especially in the presence of high TgAb concentrations.
Finally, to put the clinical value of new alternative Tg tests into critical perspective in TgAb+ patients, we should consider the positive (PPV) and negative predictive value (NPV) of these tests, which take the prevalence of recurrent/persistent disease into account, in addition to test sensitivity and specificity (24). For the clinical AWD rate of 33% in our study, which is appropriate, given the higher recurrence rates reported for TgAb+ patients (25, 26), a detectable Tg by MS in TgAb+ patients will have a PPV and NPV of approximately 65% and approximately 78%, respectively, whereas the corresponding figures for the Tg-RIA are approximately 38% and approximately 69%. In the average thyroid cancer patient population, which is primarily TgAb−, the prevalence rates of recurrent disease might be closer to 10%, resulting in predicted PPVs and NPVs of approximately 30% and approximately 95% for Tg-MS and 11% and 92% for Tg-RIA. The comparative figures for the study's Tg-IAs range from 12% to 25% for PPV and from 92% to 96% for NPV.
Acknowledgments
We thank Mary Finseth, Rebecca Wortman, and Kristen E. Hogarth for their efforts in coordinating and performing the automated immunoassays testing for this study. The UK-RIA laboratory contributions of Martin Roch and Jamie Weaver and financial support of Queen Elizabeth Hospital Birmingham Charities is gratefully acknowledged. We also thank Roche Diagnostics and Thermo Scientific for their generous gift of Tg assay reagents for the purpose of conducting this research study; these reagents are not currently commercially available in the United States.
This work was partially supported by National Institutes of Health Grants DK035816 and CA160034 (to the University of Washington).
Disclosure Summary: A.N.H. is a named inventor on a patent for antithyroglobulin peptide antibodies and received grant funding from Waters and instrument support from Waters and Thermo. B.C.N., S.K.G.G., B.G.C.L., M.R.C., P.M.C., C.A.S., A.F.T., and A.A.-S. have nothing to disclose.
Footnotes
- AFD
- alive, free of disease
- AWD
- alive with disease
- AWDD
- alive with distant disease
- AWLD
- alive with locoregional disease
- FS
- functional sensitivity
- NPV
- negative predictive value
- PPV
- positive predictive value
- ROC
- receiver-operator characteristic
- Tg
- thyroglobulin
- TgAb+
- TgAb-positive
- Tg−
- Tg negative
- TgAb
- thyroglobulin autoantibody
- Tg-IA
- Tg immunometric assay
- Tg-MS
- mass spectrometry-based Tg quantification
- Tg-RIA
- Tg RIA
- UK-RIA
- Tg-RIA performed at Queen Elizabeth Hospital
- USC-RIA
- Tg-RIA performed at the University of Southern California.
References
- 1. Spencer CA, Takeuchi M, Kazarosyan M, et al. Serum thyroglobulin autoantibodies: prevalence, influence on serum thyroglobulin measurement, and prognostic significance in patients with differentiated thyroid carcinoma. J Clin Endocrinol Metab. 1998;83:1121–1127. [DOI] [PubMed] [Google Scholar]
- 2. Spencer C, Fatemi S. Thyroglobulin antibody (TgAb) methods—strengths, pitfalls and clinical utility for monitoring TgAb-positive patients with differentiated thyroid cancer. Best Pract Res Clin Endocrinol Metab. 2013;27:701–712. [DOI] [PubMed] [Google Scholar]
- 3. Ringel MD, Nabhan F. Approach to follow-up of the patient with differentiated thyroid cancer and positive anti-thyroglobulin antibodies. J Clin Endocrinol Metab. 2013;98:3104–3110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Okosieme OE, Evans C, Moss L, Parkes AB, Premawardhana LD, Lazarus JH. Thyroglobulin antibodies in serum of patients with differentiated thyroid cancer: relationship between epitope specificities and thyroglobulin recovery. Clin Chem. 2005;51:729–734. [DOI] [PubMed] [Google Scholar]
- 5. Schneider AB, Pervos R. Radioimmunoassay of human thyroglobulin: effect of antithyroglobulin autoantibodies. J Clin Endocrinol Metab. 1978;47:126–137. [DOI] [PubMed] [Google Scholar]
- 6. Rosario PW, Maia FF, Fagundes TA, Vasconcelos FP, Cardoso LD, Purisch S. Antithyroglobulin antibodies in patients with differentiated thyroid carcinoma: methods of detection, interference with serum thyroglobulin measurement and clinical significance. Arq Bras Endocrinol Metab. 2004;48:487–492. [DOI] [PubMed] [Google Scholar]
- 7. Gorges R, Maniecki M, Jentzen W, et al. Development and clinical impact of thyroglobulin antibodies in patients with differentiated thyroid carcinoma during the first 3 years after thyroidectomy. Eur J Endocrinol. 2005;153:49–55. [DOI] [PubMed] [Google Scholar]
- 8. Verburg FA, Hartmann D, Grelle I, Giovanella L, Buck AK, Reiners C. Relationship between antithyroglobulin autoantibodies and thyroglobulin recovery rates using different thyroglobulin concentrations in the recovery buffer. Horm Metab Res. 2013;45:728–735. [DOI] [PubMed] [Google Scholar]
- 9. Verburg FA, Luster M, Cupini C, et al. Implications of thyroglobulin antibody positivity in patients with differentiated thyroid cancer: a clinical position statement. Thyroid. 2013;23:1211–1225. [DOI] [PubMed] [Google Scholar]
- 10. Spencer C, Petrovic I, Fatemi S, LoPresti J. Serum thyroglobulin (Tg) monitoring of patients with differentiated thyroid cancer using sensitive (second-generation) immunometric assays can be disrupted by false-negative and false-positive serum thyroglobulin autoantibody misclassifications. J Clin Endocrinol Metab. 2014;99:4589–4599. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Spencer C, Petrovic I, Fatemi S. Current thyroglobulin autoantibody (TgAb) assays often fail to detect interfering TgAb that can result in the reporting of falsely low/undetectable serum Tg IMA values for patients with differentiated thyroid cancer. J Clin Endocrinol Metab. 2011;96:1283–1291. [DOI] [PubMed] [Google Scholar]
- 12. Spencer CA. Clinical review: Clinical utility of thyroglobulin antibody (TgAb) measurements for patients with differentiated thyroid cancers (DTC). J Clin Endocrinol Metab. 2011;96:3615–3627. [DOI] [PubMed] [Google Scholar]
- 13. Hoofnagle AN, Becker JO, Wener MH, Heinecke JW. Quantification of thyroglobulin, a low-abundance serum protein, by immunoaffinity peptide enrichment and tandem mass spectrometry. Clin Chem. 2008;54:1796–1804. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Kushnir MM, Rockwood AL, Roberts WL, Abraham D, Hoofnagle AN, Meikle AW. Measurement of thyroglobulin by liquid chromatography-tandem mass spectrometry in serum and plasma in the presence of antithyroglobulin autoantibodies. Clin Chem. 2013;59:982–990. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Clarke NJ, Zhang Y, Reitz RE. A novel mass spectrometry-based assay for the accurate measurement of thyroglobulin from patient samples containing antithyroglobulin autoantibodies. J Investig Med. 2012;60:1157–1163. [DOI] [PubMed] [Google Scholar]
- 16. Netzel BC, Grebe SK, Algeciras-Schimnich A. Usefulness of a thyroglobulin liquid chromatography-tandem mass spectrometry assay for evaluation of suspected heterophile interference. Clin Chem. 2014;60:1016–1018. [DOI] [PubMed] [Google Scholar]
- 17. Giovanella L, Feldt-Rasmussen U, Verburg FA, Grebe SK, Plebani M, Clark PM. Thyroglobulin measurement by highly sensitive assays: focus on laboratory challenges [published online October 30, 2014]. Clin Chem Lab Med. doi:10.1515/cclm-2014-0813. [DOI] [PubMed] [Google Scholar]
- 18. Grebe SKG. Soluble thyroid tumor markers old and new challenges and potential solutions. N Z Inst Med Lab Sci J. 2013;67:76–87. [Google Scholar]
- 19. Hoofnagle AN, Wener MH. The fundamental flaws of immunoassays and potential solutions using tandem mass spectrometry. J Immunol Methods. 2009;347:3–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Hoofnagle AN, Roth MY. Clinical review: improving the measurement of serum thyroglobulin with mass spectrometry. J Clin Endocrinol Metab. 2013;98:1343–1352. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Crane MS, Strachan MW, Toft AD, Beckett GJ. Discordance in thyroglobulin measurements by radioimmunoassay and immunometric assay: a useful means of identifying thyroglobulin assay interference. Ann Clin Biochem. 2013;50:421–432. [DOI] [PubMed] [Google Scholar]
- 22. Marquet PY, Daver A, Sapin R, et al. Highly sensitive immunoradiometric assay for serum thyroglobulin with minimal interference from autoantibodies. Clin Chem. 1996;42:258–262. [PubMed] [Google Scholar]
- 23. Schlumberger M, Hitzel A, Toubert ME, et al. Comparison of seven serum thyroglobulin assays in the follow-up of papillary and follicular thyroid cancer patients. J Clin Endocrinol Metab. 2007;92:2487–2495. [DOI] [PubMed] [Google Scholar]
- 24. Altman DG, Bland JM. Diagnostic tests 2: predictive values. BMJ. 1994;309:102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Chung JK, Park YJ, Kim TY, et al. Clinical significance of elevated level of serum antithyroglobulin antibody in patients with differentiated thyroid cancer after thyroid ablation. Clin Endocrinol (Oxf). 2002;57:215–221. [DOI] [PubMed] [Google Scholar]
- 26. Durante C, Tognini S, Montesano T, et al. Clinical aggressiveness and long-term outcome in patients with papillary thyroid cancer and circulating anti-thyroglobulin autoantibodies. Thyroid. 2014;24:1139–1145. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Edge S, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A. AJCC Cancer Staging Manual. 7th ed New York: Springer; 2010. [Google Scholar]