| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Internal Medicine (R.P.P., H.v.T., E.K., T.J.V.), Erasmus University Medical Center, 3015 GE Rotterdam, The Netherlands; and Department of Intensive Care Medicine (P.J.W., G.V.d.B.), Catholic University of Leuven, B-3000 Leuven, Belgium
Address all correspondence and requests for reprints to: Greet Van den Berghe, M.D., Ph.D., Department of Intensive Care Medicine, Catholic University of Leuven, B-3000 Leuven, Belgium. E-mail: greta.vandenberghe{at}med.kuleuven.ac.be.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Results: Insulin treatment did not affect thyroid parameters. On d 1, rT3 was higher and T3/rT3 was lower in nonsurvivors as compared with survivors (P = 0.001). Odds ratio for survival of the highest vs. the lowest quartile was 0.3 for rT3 and 2.9 for T3/rT3. TSH, T4, and T3 were lower in nonsurvivors from d 5 until LD (P < 0.001). TSH, T4, T3, and T3/rT3 increased over time in survivors, but decreased or remained unaltered in nonsurvivors. Liver D1 activity was positively correlated with LD serum T3/rT3 (R = 0.83, P < 0.001) and negatively correlated with rT3 (R = –0.69, P < 0.001). Both liver and skeletal muscle D3 activity were positively correlated with LD serum rT3 (R = 0.32, P = 0.02 and R = 0.31, P = 0.03).
Conclusion: In critically ill patients who required more than 5 d of intensive care, rT3 and T3/rT3 were already prognostic for survival on d 1. On d 5, T4, T3, but also TSH levels are higher in patients who will survive. Serum rT3 and T3/rT3 were correlated with postmortem tissue deiodinase activities.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The normal or low plasma TSH levels in critically ill patients, despite decreased T4 and T3 levels, suggest a major change in setpoint within the HPT axis. Indeed, a postmortem study of patients with prolonged illness showed a decreased TRH mRNA expression in the hypothalamic paraventricular nucleus, that was correlated with decreased serum TSH and T3 (13). The combination of decreased serum T3 and increased serum rT3 levels suggests that also major changes in the peripheral metabolism of thyroid hormone occur. Reduced activation by type I deiodinase (D1) and increased inactivation by D3 have been shown in postmortem liver and skeletal muscle tissues of nonthyroidal illness patients (1, 14). An altered transport of thyroid hormone into D1- and D3-expressing tissues may also play a role (15).
Whether the reduction in serum T3 is a beneficial adaptation resulting in a decreased metabolic rate and a protection against hypercatabolism or whether it is a maladaptation contributing to a worsening of the disease is still a controversial issue (2, 3, 16). So far, it has not been clearly demonstrated that substitution of critically ill patients with thyroid hormone has a positive effect on clinical outcome (17, 18, 19). Intervention with hypothalamic-releasing factors, which restores pulsatile pituitary hormone secretion, normalizes peripheral hormone levels, and keeps the negative feedback loop intact, might be a more successful approach (20, 21, 22).
In a recent study involving 1548 intensive care unit (ICU) patients, it was shown that maintaining normoglycemia during critical illness using intensive insulin treatment reduced ICU mortality by 43% and hospital mortality by 34% (23). Furthermore, morbidity was reduced, resulting in a decreased need for prolonged mechanical ventilation, dialysis, red blood cell transfusion, antibiotic therapy, and intensive care. Previously it was shown that control of hyperglycemia in patients with either type I or type II diabetes is associated with normalization of thyroid hormone concentrations (24, 25, 26). Therefore, the metabolic effects and/or the clinical benefits of insulin may affect thyroid hormone secretion and metabolism during critical illness.
In this study, we investigated the effect of intensive insulin therapy on serum thyroid hormone levels in prolonged critically ill patients who received intensive care therapy for at least 5 d. Furthermore, we analyzed deiodinase activities in tissues of the patients who died in the ICU and investigated correlation with serum thyroid hormone levels during intensive care.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This study was part of a large randomized controlled study on intensive insulin therapy in ICU patients (n = 1548), of which the major clinical outcomes have been published in detail elsewhere (23). All mechanically ventilated adult patients were eligible for inclusion in this trial after informed consent from the closest family member. On admission, patients were randomly assigned to either strict normalization of blood glucose (80–110 mg/dl) with intensive insulin therapy or the conventional approach, in which insulin infusion is initiated only when blood glucose exceeds 215 mg/dl and which resulted in an average blood glucose of 150–160 mg/dl. The study protocol had been approved by the Ethical Review Board of the Catholic University of Leuven School of Medicine.
For the current analysis, all patients with an ICU stay of more than 5 d were included (n = 451). Table 1
describes the baseline characteristics of the two treatment groups. Blood samples were obtained at 0600 h on d 1, 5, 15, and on the last day (LD) of intensive care. Seventy-one patients died in the ICU, and liver and skeletal muscle (rectus abdominis) biopsies were available from 50 patients. Biopsies were taken within minutes after death. Sixty-two patients had been treated with thyroid hormone at some point during the course of their critical illness. Treatment was initiated when they had a serum T4 concentration below 50 nmol/liter in the face of a normal T4-binding globulin level and concomitantly clinical symptoms of hypothyroidism, defined as coma or central nervous system suppression, failure to wean from the ventilator, or hemodynamic instability, which were unexplained and resistant to conventional supportive therapy. In these cases, thyroid hormone treatment consisted of an iv bolus of 150 µg T4 daily plus 0.6 µg T3 per kilogram of body weight per 24 h as a continuous iv infusion.
|
The care of patients in the ICU often comprises infusion of heparin, either systemically or locally to prevent clotting of vascular access, which substantially interferes with the assay used to quantify free concentrations of thyroid hormone (27). Therefore, we refrained from measuring serum free T4 and free T3 in this study. Serum total T4, total T3, and TSH were measured by chemoluminescence assays (Vitros ECi Immunodiagnostic System; Ortho-Clinical Diagnostics, Amersham, UK). rT3 was measured by RIA as previously described (28). Within assay coefficients of variation amounted to 4% for TSH, 2% for T4, 2% for T3, and 3–4% for rT3. Normal values for TSH, T4, T3, and rT3 were determined in 270 healthy individuals. Mean ± 2 SD was used as the normal range for T4, T3, and rT3, whereas the 95% confidence interval was used for TSH.
Deiodinase activities
Human liver and skeletal muscle samples were homogenized on ice in 10 volumes of PE buffer (0.1 M phosphate, 2 mM EDTA, pH 7.2) using a Polytron (Kinematica AG, Lucerne, Switzerland). Homogenates were snap-frozen in aliquots and stored at –80 C until further analysis. Protein concentration was measured with the Bio-Rad Protein Assay (Bio-Rad, Veenendaal, The Netherlands) using BSA as the standard following the manufacturer’s instructions.
Liver D1 activities were determined as described earlier (14) by duplicate incubations of homogenates (10 µg protein) for 30 min at 37 C with 0.1 µM [3',5'-125I]rT3 (100,000 cpm) in a final volume of 0.1 ml PED10 buffer (PE plus 10 mM DTT). Skeletal muscle D2 activities were assayed as earlier described (14) by duplicate incubation of 200 µg of homogenate protein for 60 min at 37 C with 1 nM [3',5'-125I]T4 (100,000 cpm) in a final volume of 0.1 ml PED25 buffer (PE plus 25 mM DTT). The incubations were carried out in the presence of 0.1 µM unlabeled T3 to prevent inner ring deiodination of the labeled T4 substrate by D3, if present, and in the absence or presence of 0.1 µM unlabeled T4, which is sufficient to saturate D2. Deiodination of labeled T4 in the absence minus that in the presence of excess unlabeled T4 represents D2 activity. The further procedure for the quantitation of 125I production was the same as described above for the D1 assay. Tissue D3 activities were measured as described earlier (14) by duplicate incubation of liver (100 µg protein) or skeletal muscle (200 µg protein) homogenate for 60 min at 37 C with 1 nM [3'-125I]T3 (200,000 cpm) in a final volume of 0.1 ml PED50 buffer (PE plus 50 mM DTT).
Statistical analysis
Data were analyzed using the statistical program SPSS 10.0.7 for Windows (SPSS, Chicago, IL). To minimize the influence of outliers, all data were analyzed using nonparametric tests. Patients who had been treated with thyroid hormone were excluded from all analyses shown, except for the correlation of serum levels with tissue deiodinase activities. Data are shown as mean ± SD or as median [interquartile range (IQR)], depending on the distribution. Differences between the groups were analyzed using Mann-Whitney U tests, correlation coefficients represent Spearman’s correlation coefficient. Statistical significance was assumed for P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The baseline characteristics of the two treatment groups are shown in Table 1
. There was no difference in age, gender, severity of disease, or need for thyroid hormone treatment between the two treatment groups, but patients in the intensive insulin treatment group stayed significantly shorter in the ICU. At none of the time points (d 1, d 5, d 15, and LD) were serum TSH and iodothyronines different between the conventional and intensive insulin treatment group, although TSH and T4 tended to be lower in the intensively treated group on the LD (P = 0.08 and P = 0.06, respectively). It is most likely that only patients who were treated with thyroid hormone before their hospitalization received thyroid hormone substitution from d 1. However, thyroid hormone substitution before admission was not documented of all patients.
Serum TSH and iodothyronines are different between survivors and nonsurvivors
Table 2 shows serum TSH and iodothyronine levels of the survivors and nonsurvivors in this study. Reference values in our lab are 0.4–4.3 µU/ml for TSH, 4.51–9.95 µg/dl for T4, 92.8–162.9 ng/dl for T3, and 9.1–22.1 ng/dl for rT3. From d 5 onward, serum TSH, T4, and T3 were substantially lower in patients who will ultimately die. Serum rT3 and T3/rT3 were already different on d 1, with rT3 being higher and T3/rT3 being lower in nonsurvivors. Already on d 1, the odds ratio for survival of the highest vs. the lowest quartile was 0.30 for rT3 and 2.9 for T3/rT3. To analyze the predictive value of these parameters, we created receiver operating characteristic (ROC) curves for mortality. These revealed that the predictive value of both rT3 and T3/rT3 on d 1 was low (area under the curve was 0.68 for rT3 and 0.65 for T3/rT3), due to the overlap between survivors and nonsurvivors, and somewhat better on the LD (area under the ROC curve was 0.76 for rT3 and 0.84 for T3/rT3). The area under the ROC curve of T3/rT3 on the LD is comparable to that of Acute Physiology and Chronic Health Evaluation (APACHE) II and of IGF binding protein-1 (29). Fig. 1 shows the ROC curve of the T3 to rT3 ratio on both d 1 and the LD.
|
|
After d 1, serum TSH, T4, and T3 increased in patients who survived, whereas there was no such pattern in patients who died (Fig. 2, A–C). There was a further increase in T4 and T3 from d 5–15 and the LD, whereas TSH remained elevated from d 5 onward compared with d 1. On the LD of intensive care, the majority of survivors had TSH and T4 levels within the normal range, whereas T3 was still decreased. Serum rT3 levels were clearly elevated both in survivors and nonsurvivors throughout their stay in the ICU. In patients who did not survive, there was a substantial increase in rT3 levels on the day they died. Exclusion of patients who received glucocorticoids or of patients who were treated with dopamine at some point during intensive care gave similar results for all serum parameters.
|
). However, other than for TSH, T4, and T3, this increase in T3/rT3 in survivors occurred only after d 5. A similar course of the T3/rT3 was seen after exclusion of patients who received glucocorticoids or dopamine during intensive care. Correlation of liver D1 and liver and skeletal muscle D3 with serum iodothyronines
As described previously, we were not able to measure any D2 activity in the skeletal muscle samples of these patients, whereas D3 was induced in both liver and skeletal muscle (14). Postmortem D1 activity showed a strong, negative correlation with LD serum rT3 (R = –0.69, P < 0.001) and a positive correlation with LD serum T3/rT3 (R = 0.83, P < 0.001) (Fig. 3), but not with T4 or T3 (Table 3). Postmortem liver and muscle D3 activities showed a significant positive correlation with serum rT3 levels (Fig. 3), but were not significantly associated with the T3 to rT3 ratio (Table 3). Furthermore, liver but not skeletal muscle D3 was negatively associated with serum T3 levels (Table 3).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
During critical illness, plasma T3 decreases and rT3 increases, and the magnitude of these changes is related to the severity of illness (2, 5, 6, 7). Intensive insulin therapy has been shown to result in a decreased morbidity and mortality in intensive care patients (23, 30), and it has been shown that control of hyperglycemia in patients with diabetes is associated with normalization of thyroid hormone concentrations (24, 25, 26). However, no difference in serum thyroid parameters was observed between patients treated with intensive and conventional insulin therapy, suggesting that neither insulin per se, nor its metabolic effects or its clinical benefits, resulted in altered thyroid hormone levels. For this study, only patients who stayed in the ICU for more than 5 d were selected. Because intensive insulin therapy also resulted in a shorter period of ICU stay (23), the possibility of a selection bias (sicker patients in the studied group receiving intensive insulin therapy) cannot be excluded. Furthermore, an effect of insulin on the nocturnal TSH peak cannot be excluded from these data, although this is unlikely because such an effect would have resulted in differences in serum T4 and T3 levels (26).
Previous studies have shown that patients with more pronounced alterations in serum thyroid parameters have a significantly higher mortality rate (1, 2, 7, 31, 32). Although all patients required more than 5 d of intensive care, rT3 and T3/rT3 on d 1 were already prognostic for survival, whereas TSH, T4, and T3 were significantly different between survivors and nonsurvivors from d 5 onward. Not only the absolute values, but also the time course was completely different between survivors and nonsurvivors. TSH, T4, and T3 increased in patients who survived, whereas there was no such rise in patients who died. T4 and T3 continued to increase from d 1 to the LD, whereas TSH did not further rise after d 5, suggesting that both T4 and T3 follow the initial increase in TSH. This is in line with previous observations suggesting that an increase in serum TSH drives the rise in T4 and marks the onset of recovery (33, 34). This driving TSH surge is likely to be mediated by TRH. Indeed, hypothalamic TRH gene expression was previously found to be positively correlated with serum TSH and T3 during prolonged critical illness (13), and infusion of TRH in the critically ill could normalize peripheral thyroid hormone levels (20, 21, 22). However, a full recovery of the HPT axis in this phase of illness, combined with the low T3 levels, would have required an elevated level of TSH. This was not observed in the current study, whereas it has been described by others (33, 34). This difference cannot be explained by the treatment with glucocorticoids or dopamine, because exclusion of all patients who were treated with these drugs gave similar results. However, it may be due to the infrequent serum analysis, because increased levels of TSH were seen in certain, but not all patients.
Serum rT3 levels were clearly elevated in both survivors and nonsurvivors throughout the intensive care, and rT3 levels did not decrease in survivors. Interestingly, in patients who did not survive, there was a substantial further rise in rT3 levels toward the LD. This might be explained by a short half-life of rT3 (around 3 h vs. around 24 h for T3) (7, 35, 36, 37), making rT3 a sensitive marker for acute changes in thyroid hormone metabolism that are caused by perimortem tissue decay.
On the LD of intensive care, the majority of patients had TSH and T4 levels back within the normal range, whereas T3 and rT3 remained outside the normal range. Thus, T3 and rT3 levels are not only the first to change in the acute phase of illness (1, 2, 4, 7), but also the last ones to recover.
Serum iodothyronine levels depend not only on the activities of iodothyronine-metabolizing enzymes, but also, among other things, on thyroid function and serum iodothyronine-binding capacity. Because of the confounding effect of variable concentrations of T4 and T4-binding proteins, the serum T3 to rT3 ratio is the parameter that most accurately reflects the result of altered peripheral thyroid hormone metabolism during critical illness. The serum T3 to rT3 ratio increased with time in survivors, whereas it did not alter or even decreased in nonsurvivors. This increase in T3/rT3 in survivors occurred only after d 5, suggesting that the peripheral metabolism recovers at a later stage than the centrally initiated TSH secretion.
A low D1 activity will result in a reduced T3 production and rT3 clearance (7). On the other hand, D1 expression is under positive control by thyroid hormone, which is thought to be mediated by T3 (7, 38). In agreement with this, liver D1 was negatively correlated with rT3 and positively with T3/rT3. An induced D3 expression enhances T3 clearance and rT3 production (7, 14), and in line with this, liver and skeletal muscle D3 were positively correlated with rT3, whereas liver D3 also showed a negative correlation with serum T3. Our data suggest an important role of liver D1, and of liver and skeletal muscle D3 in altering thyroid hormone levels in critically ill patients. Especially the very strong correlation between liver D1 and the serum T3 to rT3 ratio is remarkable, because serum samples were obtained at 0600 h on the day the patient died, and deiodinase activities were measured in postmortem biopsies.
It should be noted that we were not able to detect any D2 in skeletal muscle samples of these patients, whereas D2 activity is present in skeletal muscle of normal subjects (7, 39). There is evidence that D2 in skeletal muscle also contributes to serum T3 production, especially in the hypothyroid situation (7, 39, 40). Therefore, it is likely that a decreased T4 to T3 conversion by skeletal muscle D2 may also contribute to the low T3 levels in critically ill patients.
In conclusion, in critically ill patients who required intensive care for more than 5 d, rT3 and T3/rT3 on d 1 were already prognostic for survival although the ROC curve revealed a rather low predictive value. From d 5 onward, TSH, T4, T3 levels, and in a later stage, T3/rT3 increased in survivors, suggesting that recovery of thyroid parameters is initiated centrally. The centrally initiated increase in TSH secretion was followed by a recovery of the peripheral metabolism. Serum rT3 and T3/rT3 were significantly correlated with postmortem deiodinase activity.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
First Published Online May 10, 2005
Abbreviations: HPT, Hypothalamus-pituitary-thyroid; ICU, intensive care unit; IQR, interquartile range; LD, last day; ROC, receiver operating characteristic.
Received March 10, 2005.
Accepted May 2, 2005.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. J. Lee, J. G. Kang, O. H. Ryu, C. S. Kim, S.-H. Ihm, M. G. Choi, H. J. Yoo, and K. S. Hong The relationship of thyroid hormone status with myocardial function in stress cardiomyopathy Eur. J. Endocrinol., May 1, 2009; 160(5): 799 - 806. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E Forestier, S Vinzio, R Sapin, J L Schlienger, and B Goichot Increased reverse triiodothyronine is associated with shorter survival in independently-living elderly: the Alsanut study Eur. J. Endocrinol., February 1, 2009; 160(2): 207 - 214. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Gereben, A. M. Zavacki, S. Ribich, B. W. Kim, S. A. Huang, W. S. Simonides, A. Zeold, and A. C. Bianco Cellular and Molecular Basis of Deiodinase-Regulated Thyroid Hormone Signaling Endocr. Rev., December 1, 2008; 29(7): 898 - 938. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Dalan, M. K.-S. Leow, and D. C.-E. Ng High iodine (substrate) turnover Graves' disease: the intriguing 'rapid responder' variant of thyrotoxicosis Ann Clin Biochem, November 1, 2008; 45(6): 612 - 615. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Debaveye, B. Ellger, L. Mebis, T. J. Visser, V. M. Darras, and G. Van den Berghe Effects of Substitution and High-Dose Thyroid Hormone Therapy on Deiodination, Sulfoconjugation, and Tissue Thyroid Hormone Levels in Prolonged Critically Ill Rabbits Endocrinology, August 1, 2008; 149(8): 4218 - 4228. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A Boelen, J Kwakkel, W M Wiersinga, and E Fliers Chronic local inflammation in mice results in decreased TRH and type 3 deiodinase mRNA expression in the hypothalamic paraventricular nucleus independently of diminished food intake J. Endocrinol., December 1, 2006; 191(3): 707 - 714. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Kohrle, F. Jakob, B. Contempre, and J. E. Dumont Selenium, the Thyroid, and the Endocrine System Endocr. Rev., December 1, 2005; 26(7): 944 - 984. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. P. Peeters, M. H. A. Kester, P. J. Wouters, E. Kaptein, H. van Toor, T. J. Visser, and G. Van den Berghe Increased Thyroxine Sulfate Levels in Critically Ill Patients as a Result of a Decreased Hepatic Type I Deiodinase Activity J. Clin. Endocrinol. Metab., December 1, 2005; 90(12): 6460 - 6465. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. P. Peeters, S. van der Geyten, P. J. Wouters, V. M. Darras, H. van Toor, E. Kaptein, T. J. Visser, and G. Van den Berghe Tissue Thyroid Hormone Levels in Critical Illness J. Clin. Endocrinol. Metab., December 1, 2005; 90(12): 6498 - 6507. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Debaveye, B. Ellger, L. Mebis, E. Van Herck, W. Coopmans, V. Darras, and G. Van den Berghe Tissue Deiodinase Activity during Prolonged Critical Illness: Effects of Exogenous Thyrotropin-Releasing Hormone and Its Combination with Growth Hormone-Releasing Peptide-2 Endocrinology, December 1, 2005; 146(12): 5604 - 5611. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
