Clinical Endocrinology (2016) 84, 305–309
LETTERS TO THE EDITOR
Standardization of FT4 and harmonization of TSH measurements – a request for input from endocrinologists and other physicians Dear editor, Given the prevalence of thyroid disorders and the subtle signs and symptoms which may accompany subclinical disease, reliable laboratory testing for serum TSH and free thyroid hormones is important for both primary care physicians and endocrinologists. The laboratory community has recognized the need for standardization of thyroid function tests to achieve comparability of measurement results between methods. This applies particularly for free T4 tests, which may be considered controversial in terms of clinical and analytical validity. But there is also variability in TSH testing – a fact which has not been emphasized in ongoing discussions regarding lowering of the upper limit of normal and/or common decision limits to start treatment for hypothyroidism. In response to this need, the International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) formed a Working Group for Standardization of Thyroid Function Tests (WG-STFT) in 2005. Today, the WG is a full Committee (C-STFT), and the authors are, respectively, chair and member of the C-STFT and past president at IFCC. We have worked over the years towards the goal of standardization of free T4 and TSH testing. Because, unlike free T4, there is no reference measurement procedure for TSH, we have proposed an alternative to standardization for this test called harmonization.1–4 Different phases of method comparison studies were performed to investigate and confirm the feasibility of standardization of FT4 and harmonization of TSH tests, but also with attention to fit-for-the-purpose analytical quality.5–7 For free T4, the variability is substantial. Standardization will change results significantly – perhaps as much as 80% at the upper limit of the normal range – for some assays. For TSH, the alterations introduced by harmonization will be milder (approximately 20%). The Committee, which comprises representatives from both laboratory medicine and in vitro diagnostic manufacturers, discussed these data with the appropriate Food and Drug Administration authorities in early 2014. They agreed that the Committee’s approach was reasonable but asked us to investigate how physicians and patients would react to the changes in the numerical results for some of the assays. To this end, we are reaching out to stakeholders to seek views on the benefits and risks arising from FT4 standardization and TSH harmonization. • Benefits: We wish to learn of the benefits that you think would be achieved if all FT4 and TSH assays gave comparable results on patient samples.
© 2015 John Wiley & Sons Ltd
• Risks: We wish to learn of the risks to patient safety and clinical outcomes that you think may arise as a consequence of a change in the numerical results for patients with thyroid disorders who are being followed, despite the appropriate similar change in the reference intervals. • Implementation: We wish to hear your views on how clinical laboratories can minimize the identified risks from FT4 standardization and TSH harmonization, and on the role that you could play yourself in your contact with the patients you treat. This call is being launched in parallel to several clinical journals and to professional organizations for doctors and laboratory specialists.
Competing interests Nothing to declare.
Financial disclosure Nothing to declare. Linda M. Thienpont*, James D. Faix†,‡ and Graham Beastall§ *Mass Spectrometry Reference Laboratory LAC, Gent University, Gent, Belgium †Clinical Chemistry and Immunology, Montefiore Medical Center, ‡Pathology, Albert Einstein School of Medicine, New York, NY, USA and §International Federation of Clinical Chemistry & Laboratory Medicine, Glasgow, UK E-mail: [email protected] doi: 10.1111/cen.12861
References 1 Thienpont, L.M., Beastall, G., Christofides, N.D. et al. (2007) Measurement of free thyroxine in laboratory medicine–proposal of measurand definition. Clinical Chemistry and Laboratory Medicine, 45, 563–564. 2 Thienpont, L.M., Beastall, G., Christofides, N.D. et al. (2007) Proposal of a candidate international conventional reference measurement procedure for free thyroxine in serum. Clinical Chemistry and Laboratory Medicine, 45, 934–936. 3 Van Houcke, S.K., Van Uytfanghe, K., Shimizu, E. et al. (2011) IFCC international conventional reference procedure for the measurement of free thyroxine in serum. Clinical Chemistry and Laboratory Medicine, 49, 1275–1281. 4 St€ ockl, D., Van Uytfanghe, K., Van Aelst, S. et al. (2014) A statistical basis for harmonization of thyroid stimulating hormone immunoassays using a robust factor analysis model. Clinical Chemistry and Laboratory Medicine, 52, 965–972.
305
306 Letters to the Editor 5 Thienpont, L.M., Van Uytfanghe, K., Beastall, G. et al. (2010) Report of the IFCC Working Group for Standardization of Thyroid Function Tests, part 1: thyroid-stimulating hormone. Clinical Chemistry, 56, 902–911. 6 Thienpont, L.M., Van Uytfanghe, K., Beastall, G. et al. (2010) Report of the IFCC Working Group for Standardization of Thyroid Function Tests, part 2: free thyroxine and free triiodothyronine. Clinical Chemistry, 56, 912–920. 7 Thienpont, L.M., Van Uytfanghe, K., Van Houcke, S. et al. (2014) A Progress report of the IFCC Committee for Standardization of Thyroid Function Tests. European Thyroid Journal, 3, 109–116.
Pancreatic fat and hepato-metabolic features in obese children with nonalcoholic fatty liver disease We have several concerns about the paper by Della Corte et al.1 ‘Non-alcoholic fatty pancreas disease and non-alcoholic fatty liver disease: more than ectopic fat’ submitted to Clinical Endocrinology on May 15 and published online on July 23. • First, we respectfully disagree that their report is the first study aiming to clarify the possible link between fatty pancreas and hepato-metabolic features that characterize obese children affected by nonalcoholic fatty liver disease (NAFLD), defined as echographic evidence of fatty liver after exclusion of other causes of hepatic steatosis. That does not hold true. In an earlier paper published online on April 21,2 we determined whether magnetic resonance imaging (MRI)-calculated pancreatic fat fraction (PFF) was greater in overweight/obese Caucasian children and adolescents with NAFLD (defined on the basis of a validated threshold for MRI hepatic fat fraction in the absence of secondary causes of hepatic steatosis) than in those without liver involvement. We also assessed the relationship between PFF and other fat depots, including liver fat, visceral and subcutaneous adipose tissue; and finally, we explored the association between pancreatic fat accumulation and b-cell dysfunction and increased risk for prediabetes. It is therefore surprising that none of the 10 co-authors admitted the pre-existing paediatric literature on the subject since submission until the date of acceptance. • Second, the method for detection of fatty pancreas makes it uncertain the interpretation of study design and findings. Fatty pancreas was diagnosed in the presence of increased echogenicity of the pancreatic parenchyma over that of the kidney. As acknowledged by authors, as the pancreas cannot be compared with the kidney in the same window ultrasound, the radiologists compared the difference between hepatic and renal echogenicity, and between hepatic and pancreas echogenicity. However, the liver may be not useful as a reference structure, because interpretation of mild fatty liver infiltration has considerable inter- or intra-observer variability on ultrasound.3 Visual judgment of pancreatic
echogenicity compared with the liver or kidney can be difficult, because normal pancreas shows equal to or slightly greater echogenicity than that of the liver or kidney.4 In addition, the different gain and focus on different image windows can make the comparison between the pancreas and kidney uncertain. Besides, it is unclear whether subjects with poor image quality of pancreas echogenicity due to severe obesity and bowel gas artefacts were excluded. Further major drawbacks of sonography include operator dependency and insensitivity with regard to small amounts of fat. According to authors, patients underwent abdominal ultrasound evaluation ‘by the same three experienced radiologist’. Regrettably, the interobserver percentage of agreement (i.e. kappa statistic) for ultrasound diagnosis of fatty pancreas was not reported. Lastly, another weakness of their study was the lack of the threshold of pancreatic fat accumulation leading to metabolic outcomes. • Third, the patients with nonalcoholic fatty pancreas disease (NAFPD) exhibited a significantly higher z-body mass index (BMI), fasting insulin, insulin-resistance (HOMA-IR) and lower insulin sensitivity index with respect to the group without fatty pancreas. Based on their findings, the authors suggested that fatty pancreas should be considered an additional factor contributing to the development of gluco-insulinemic disarray in obese patients with NAFLD. However, we respectfully suggest that their report contains insufficient evidence to support that conclusion. No multivariate analysis was performed to adjust for z-BMI and other confounding covariates including abdominal fat depots. As such, the synergistic effect of NAFPD and NAFLD on glucose metabolism and b-cell function is a valid hypothesis about which their data permit no more than speculation. To further complicate matters, no data were provided in children without NAFLD. • Finally, a cautionary note on the positive association between fatty pancreas and severity of liver injury is suggested. Recently, Patel et al.5 demonstrated that MRI-determined pancreatic fat content was significantly higher in adult patients without evidence of fibrosis on liver biopsy than in those with fibrosis. The mean MRI fat fraction of the pancreas was 112% in 19 patients with no fibrosis, 58% in 13 patients with mild fibrosis and 69% in 11 patients with advanced fibrosis.
Competing interests/financial disclosure Nothing to declare. Lucia Pacifico and Claudio Chiesa Institute of Translational Pharmacology, National Research Council, Rome, Italy E-mail: [email protected] doi: 10.1111/cen.12870
© 2015 John Wiley & Sons Ltd Clinical Endocrinology (2016), 84, 305–309
LETTERS TO THE EDITOR
Standardization of FT4 and harmonization of TSH measurements – a request for input from endocrinologists and other physicians Dear editor, Given the prevalence of thyroid disorders and the subtle signs and symptoms which may accompany subclinical disease, reliable laboratory testing for serum TSH and free thyroid hormones is important for both primary care physicians and endocrinologists. The laboratory community has recognized the need for standardization of thyroid function tests to achieve comparability of measurement results between methods. This applies particularly for free T4 tests, which may be considered controversial in terms of clinical and analytical validity. But there is also variability in TSH testing – a fact which has not been emphasized in ongoing discussions regarding lowering of the upper limit of normal and/or common decision limits to start treatment for hypothyroidism. In response to this need, the International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) formed a Working Group for Standardization of Thyroid Function Tests (WG-STFT) in 2005. Today, the WG is a full Committee (C-STFT), and the authors are, respectively, chair and member of the C-STFT and past president at IFCC. We have worked over the years towards the goal of standardization of free T4 and TSH testing. Because, unlike free T4, there is no reference measurement procedure for TSH, we have proposed an alternative to standardization for this test called harmonization.1–4 Different phases of method comparison studies were performed to investigate and confirm the feasibility of standardization of FT4 and harmonization of TSH tests, but also with attention to fit-for-the-purpose analytical quality.5–7 For free T4, the variability is substantial. Standardization will change results significantly – perhaps as much as 80% at the upper limit of the normal range – for some assays. For TSH, the alterations introduced by harmonization will be milder (approximately 20%). The Committee, which comprises representatives from both laboratory medicine and in vitro diagnostic manufacturers, discussed these data with the appropriate Food and Drug Administration authorities in early 2014. They agreed that the Committee’s approach was reasonable but asked us to investigate how physicians and patients would react to the changes in the numerical results for some of the assays. To this end, we are reaching out to stakeholders to seek views on the benefits and risks arising from FT4 standardization and TSH harmonization. • Benefits: We wish to learn of the benefits that you think would be achieved if all FT4 and TSH assays gave comparable results on patient samples.
© 2015 John Wiley & Sons Ltd
• Risks: We wish to learn of the risks to patient safety and clinical outcomes that you think may arise as a consequence of a change in the numerical results for patients with thyroid disorders who are being followed, despite the appropriate similar change in the reference intervals. • Implementation: We wish to hear your views on how clinical laboratories can minimize the identified risks from FT4 standardization and TSH harmonization, and on the role that you could play yourself in your contact with the patients you treat. This call is being launched in parallel to several clinical journals and to professional organizations for doctors and laboratory specialists.
Competing interests Nothing to declare.
Financial disclosure Nothing to declare. Linda M. Thienpont*, James D. Faix†,‡ and Graham Beastall§ *Mass Spectrometry Reference Laboratory LAC, Gent University, Gent, Belgium †Clinical Chemistry and Immunology, Montefiore Medical Center, ‡Pathology, Albert Einstein School of Medicine, New York, NY, USA and §International Federation of Clinical Chemistry & Laboratory Medicine, Glasgow, UK E-mail: [email protected] doi: 10.1111/cen.12861
References 1 Thienpont, L.M., Beastall, G., Christofides, N.D. et al. (2007) Measurement of free thyroxine in laboratory medicine–proposal of measurand definition. Clinical Chemistry and Laboratory Medicine, 45, 563–564. 2 Thienpont, L.M., Beastall, G., Christofides, N.D. et al. (2007) Proposal of a candidate international conventional reference measurement procedure for free thyroxine in serum. Clinical Chemistry and Laboratory Medicine, 45, 934–936. 3 Van Houcke, S.K., Van Uytfanghe, K., Shimizu, E. et al. (2011) IFCC international conventional reference procedure for the measurement of free thyroxine in serum. Clinical Chemistry and Laboratory Medicine, 49, 1275–1281. 4 St€ ockl, D., Van Uytfanghe, K., Van Aelst, S. et al. (2014) A statistical basis for harmonization of thyroid stimulating hormone immunoassays using a robust factor analysis model. Clinical Chemistry and Laboratory Medicine, 52, 965–972.
305
306 Letters to the Editor 5 Thienpont, L.M., Van Uytfanghe, K., Beastall, G. et al. (2010) Report of the IFCC Working Group for Standardization of Thyroid Function Tests, part 1: thyroid-stimulating hormone. Clinical Chemistry, 56, 902–911. 6 Thienpont, L.M., Van Uytfanghe, K., Beastall, G. et al. (2010) Report of the IFCC Working Group for Standardization of Thyroid Function Tests, part 2: free thyroxine and free triiodothyronine. Clinical Chemistry, 56, 912–920. 7 Thienpont, L.M., Van Uytfanghe, K., Van Houcke, S. et al. (2014) A Progress report of the IFCC Committee for Standardization of Thyroid Function Tests. European Thyroid Journal, 3, 109–116.
Pancreatic fat and hepato-metabolic features in obese children with nonalcoholic fatty liver disease We have several concerns about the paper by Della Corte et al.1 ‘Non-alcoholic fatty pancreas disease and non-alcoholic fatty liver disease: more than ectopic fat’ submitted to Clinical Endocrinology on May 15 and published online on July 23. • First, we respectfully disagree that their report is the first study aiming to clarify the possible link between fatty pancreas and hepato-metabolic features that characterize obese children affected by nonalcoholic fatty liver disease (NAFLD), defined as echographic evidence of fatty liver after exclusion of other causes of hepatic steatosis. That does not hold true. In an earlier paper published online on April 21,2 we determined whether magnetic resonance imaging (MRI)-calculated pancreatic fat fraction (PFF) was greater in overweight/obese Caucasian children and adolescents with NAFLD (defined on the basis of a validated threshold for MRI hepatic fat fraction in the absence of secondary causes of hepatic steatosis) than in those without liver involvement. We also assessed the relationship between PFF and other fat depots, including liver fat, visceral and subcutaneous adipose tissue; and finally, we explored the association between pancreatic fat accumulation and b-cell dysfunction and increased risk for prediabetes. It is therefore surprising that none of the 10 co-authors admitted the pre-existing paediatric literature on the subject since submission until the date of acceptance. • Second, the method for detection of fatty pancreas makes it uncertain the interpretation of study design and findings. Fatty pancreas was diagnosed in the presence of increased echogenicity of the pancreatic parenchyma over that of the kidney. As acknowledged by authors, as the pancreas cannot be compared with the kidney in the same window ultrasound, the radiologists compared the difference between hepatic and renal echogenicity, and between hepatic and pancreas echogenicity. However, the liver may be not useful as a reference structure, because interpretation of mild fatty liver infiltration has considerable inter- or intra-observer variability on ultrasound.3 Visual judgment of pancreatic
echogenicity compared with the liver or kidney can be difficult, because normal pancreas shows equal to or slightly greater echogenicity than that of the liver or kidney.4 In addition, the different gain and focus on different image windows can make the comparison between the pancreas and kidney uncertain. Besides, it is unclear whether subjects with poor image quality of pancreas echogenicity due to severe obesity and bowel gas artefacts were excluded. Further major drawbacks of sonography include operator dependency and insensitivity with regard to small amounts of fat. According to authors, patients underwent abdominal ultrasound evaluation ‘by the same three experienced radiologist’. Regrettably, the interobserver percentage of agreement (i.e. kappa statistic) for ultrasound diagnosis of fatty pancreas was not reported. Lastly, another weakness of their study was the lack of the threshold of pancreatic fat accumulation leading to metabolic outcomes. • Third, the patients with nonalcoholic fatty pancreas disease (NAFPD) exhibited a significantly higher z-body mass index (BMI), fasting insulin, insulin-resistance (HOMA-IR) and lower insulin sensitivity index with respect to the group without fatty pancreas. Based on their findings, the authors suggested that fatty pancreas should be considered an additional factor contributing to the development of gluco-insulinemic disarray in obese patients with NAFLD. However, we respectfully suggest that their report contains insufficient evidence to support that conclusion. No multivariate analysis was performed to adjust for z-BMI and other confounding covariates including abdominal fat depots. As such, the synergistic effect of NAFPD and NAFLD on glucose metabolism and b-cell function is a valid hypothesis about which their data permit no more than speculation. To further complicate matters, no data were provided in children without NAFLD. • Finally, a cautionary note on the positive association between fatty pancreas and severity of liver injury is suggested. Recently, Patel et al.5 demonstrated that MRI-determined pancreatic fat content was significantly higher in adult patients without evidence of fibrosis on liver biopsy than in those with fibrosis. The mean MRI fat fraction of the pancreas was 112% in 19 patients with no fibrosis, 58% in 13 patients with mild fibrosis and 69% in 11 patients with advanced fibrosis.
Competing interests/financial disclosure Nothing to declare. Lucia Pacifico and Claudio Chiesa Institute of Translational Pharmacology, National Research Council, Rome, Italy E-mail: [email protected] doi: 10.1111/cen.12870
© 2015 John Wiley & Sons Ltd Clinical Endocrinology (2016), 84, 305–309
