Analytical Variation Between Two Different TSH Reagents from the Same Manufacturer

Abstract

Thyroid stimulating hormone (TSH) immunoassays are known for giving varying results based on the platform of testing and the generation of kit used. It is generally expected that the results should not vary to affect clinical diagnosis and management. We aimed to perform method comparison study between two TSH assays by the same manufacturer Siemens Healthineers. Results show that there is a large proportional error between the assays with a bias of -3.71mIu/L indicating that TSH assay gives higher values for TSH for the same patient as measured against the TSH3-Ultra kit. This can affect interpretation of results leading to false increase in patients categorized under hypothyroidism and subclinical hypothyroidism. We strongly suggest, to prevent errors in clinical evaluation of a patient with thyroid dysfunction, validation of the performance of the assay and method comparison should be performed in-house.

Introduction

Laboratory measurement of serum Thyroid stimulating hormone (TSH) is an essential tool for diagnosing and managing various thyroid disorders. However, it is well known that immunoassays continue to be an important source of variability in patients’ results. [1] The earlier TSH tests which were competitive immunoassays used polyclonal antibodies; however, the analytical sensitivity of these assays was not sufficient to differentiate hyperthyroid patients with suppressed TSH from euthyroid individuals. Therefore, non-competitive immunoassays were introduced by using monoclonal antibodies. These second- and third-generation assays detect TSH with a degree of reproducibility of approximately 0.1 and 0.01 mIU/L, respectively. [2] The majority of currently available TSH immunoassays are capable of third-generation performance. Nevertheless, laboratory medicine professionals have noted that in practice the assays do not provide equivalent/similar results. Thus, the International Federation of Clinical Chemistry (IFCC)Working Group for Standardization of Thyroid Function undertook a TSH harmonization exercise where 14 manufacturers took part in the effort to standardize the TSH results and make efforts to reduce the inter-assay variability. [3]

It has been noticed that sometimes manufacturers have multiple kits for the same assay with the newer ones claimed to show better analytical performance. On the other hand, physicians may not be aware as to how it affects the interpretation of results. Thus, the aim of this study was to evaluate the performance of two such TSH assays provided by Siemens Healthineers (Erlangen, Germany), which are used in the laboratory in the diagnosis of thyroid disorders.

Materials and Methods

Siemens Healthineers (Erlangen, Germany) has two assays for the measurement of TSH3-Ultra (TSH-U) and TSH for use on their systems Advia Centaur, Advia Centaur XP and Advia Centaur XPT systems. The assay range for the kits are 0.008–150 mIU/L and 0.01–150 mIU/L respectively, making them both 3rd generation kits (pack inserts). A total of 25 randomly selected samples were analyzed in triplicate on Advia Centaur XP platform using Siemens’s standard protocol. The internal controls (Lyphochek Immunoassay Plus controls, Bio-Rad, Hercules, California) were run as per protocol prescribed by NABL112-ISO15189. [4] The limit of acceptance for daily internal controls of the provided three levels was CV < 5%. Statistical analysis was done using Analyse-it (Version 5.65.7).

Results

Range of patient results: On TSH-U the patient’s TSH results ranged from 0.01–63.39 mIU/L and TSH: 0.02–41.65-mIU/L. Passing–Bablok regression calculated an intercept of 0.23 but slope of 1.41 (95% CI 1.28–1.52) suggestive of small constant difference but a large proportional difference (Fig. 1). Also, Bland Altman Plot does show a bias of −3.77 mIU/L with wide limits of agreement (Fig. 2).

Fig. 1.

Passing Bablok regression line between TSH-U and TSH

Fig. 2.

Bland Altman plot between TSH-U and TSH (in absolute units)

Discussion

Clinical decisions are often based on guidelines that use a fixed lab test result. The upper limit of euthyroidism with first-generation TSH assays was approximately 10 mIU/L, but with the introduction of second- and third-generation assays it fell to approximately 5 mIU/L. (2). In pregnant women, the upper limit of normal for TSH is kept at 5mIU/L. Moreover, for subclinical hypothyroid women undergoing IVF or intracytoplasmic sperm injection (ICSI), the goal of treatment is to achieve a TSH concentration < 2.5 mIU/L. [5] Every manufacturer has determined their upper reference ranges which is more or less around 5 mIU/L but not uniform. The reference range for the kits are as follows: TSH3-U 0.55–4.78 mIU/L and TSH kit 0.35–5.50 mIU/L. The TSH3-ultra kit is the newer kit with higher sensitivity and was submitted for evaluation and standardization to the IFCC Working Group for Standardization of Thyroid Function and thus has been considered as the reference method for comparison. [3]. The results of the study indicate that there is significant bias at the upper limit of reference range between the older TSH and the TSH-U assays and correlation between them were poor. This difference increases proportionally as the value of TSH increases. Polyclonal antibodies often contain multiple epitopes. The older TSH kit uses a two-site sandwich immunoassay where the first antibody is a monoclonal mouse anti-TSH antibody labelled with acridinium ester. The second antibody, in the solid phase, is a polyclonal sheep anti-TSH antibody which is covalently coupled to paramagnetic particles. On the other hand, TSH-U employs anti-FITC monoclonal antibody covalently bound to paramagnetic particles, an FITC-labelled anti-TSH capture monoclonal antibody, and a tracer of acridinium ester and an anti-TSH mAb antibody [pack inserts]. Therefore, it is plausible that both the kits don’t produce equivalent results.

Our study showed that the older TSH kit gives higher results as compared to the TSH-U kit. This is especially important for patients who may get tested in different laboratories during the course of their treatment. The clinician may consider that any result outside the provided reference range [which is similar in both assays] as clinically important. This will result in the possible misclassification of patients. These patients may be asymptomatic or have nonspecific symptoms found in the general population without any thyroid dysfunction and may be put under medication as a result of diagnostic misclassification. Patients diagnosed with subclinical hypothyroidism or hypothyroidism on treatment can be classified as non-responders and the dosage of levothyroxine might be increased leading to increased risk of side effects. For a pregnant patient, the high results obtained when tested by the older TSH kit can possibly be labelled as hypothyroidism leading to unnecessary treatment which may have serious consequence. It may be further noted that this is a difficult problem to detect in routine EQA programs, as the assay performance is compared relative to other laboratories using the same method, same instrument, and same kit i.e. respective peer group.

Conclusion

The Siemens TSH assay gives higher values as compared to the Siemens TSH-U assay. Thus, measurement by the TSH assay is inappropriate in the presence of the standardized TSH-U assay. The marked variations of results between different analytical platform has been highlighted in another study [6] but to the best of our knowledge this is the first time that such variation is being reported on the same platform on reagents from the same manufacturer. The probable limitation of the study is that the samples tested didn’t cover the complete analytical measurement range. Nevertheless, this initial observation is clinically relevant. As laboratory professionals, we should be responsible to use newer developed assays, verify the performance of the assays in-house to provide an acceptable level of diagnosis and monitoring for patients. We must also encourage manufacturers to support only properly standardized assays with applicable reference intervals.