INTRODUCTION/TERMINOLOGY
In the absence of hypothalamic, pituitary or thyroid diseases, systemic illnesses have multiple effects on thyroid hormone metabolism and on serum thyroid hormone concentrations (1-3). These changes are primarily related to the severity (2) (Figure ), and perhaps to the chronicity of illnesses (4), rather than to the specific disease states. Changes related to chronicity may be secondary to worsening malnutrition and/or catabolism which result in progressive declines in thyroid binding protein concentrations.


Figure : Relationship of serum thyroid hormone levels to the severity of nonthyroidal illnesses. Adapted from (3).

Most frequently, serum levels of total T3 are reduced leading to the term “the low T3 syndrome” or “the low T3 state of nonthyroidal illness”. Less frequently, serum total and free T4 values and TSH concentrations are increased or decreased. Some have termed patients with low total T4 and T3 levels as “the low T3-low T4 state” of nonthyroidal illness. Many authors believe that these patients are euthyroid, hence the label “euthyroid sick syndrome”, although this area remains controversial (5, 6). The term “nonthyroidal illness syndrome” (NTIS) encompasses the transient changes in serum thyroid hormone levels as well as the alterations in thyroid hormone metabolism induced by systemic illnesses in patients without concurrent hypothalamic, pituitary or thyroid diseases, and does not imply thyroid hormone status.

GOALS
The primary goals in clinical practice are to 1) understand the pathophysiology of changes in thyroid hormone metabolism and serum thyroid hormone alterations in NTIS and the potential clinical significance of these alterations, 2) differentiate changes in serum thyroid hormone levels due to NTIS from those due to concurrent hyper- or hypothyroidism, and 3) recognize effects of NTIS on serum thyroid hormone levels in patients with concurrent hyperthyroidism or hypothyroidism. In addition, a number of assumptions have been made regarding the thyroid hormone status of patients with NTIS and the potential beneficial or detrimental effects of thyroid hormone replacement therapy in these patients. These issues will be summarized.

EPIDEMIOLOGY OF NTIS
Serum total T3 concentrations are frequently reduced even with mild nonthyroidal disorders such as caloric deprivation, with the highest frequency and lowest values occurring in patients with the most severe illnesses (1, 2). Serum total T3 values return to normal only after complete recovery of the nonthyroidal illness and of the nutritional deficiency.
In the general population, altered free T4 index values are as frequently due to NTIS as to hyper- or hypothyroidism (1). Free T4 index values were transiently increased in 0.2 to 0.9% due to NTIS and in 0.3 to 0.5% due to hyperthyroidism. Transiently increased free T4 estimates are usually associated with mild nonthyroidal illnesses and acute psychiatric disorders (1). Free T4 index values were transiently reduced in 0.2 to 1.1% due to NTIS and in 0.6 to 1.1% of patients due to primary hypothyroidism (1). Transiently reduced free T4 estimates (indexes and immunoassays) occur most commonly with moderate to severe nonthyroidal illnesses, with the lowest values occurring in critically ill patients with the highest mortality rates (1). In NTIS patients with altered free T4 estimates, serum TSH values are usually within thereference range of values for euthyroid subjects facilitating differentiation from primary hypothyroidism or hyperthyroidism (1).
In hospitalized patients, serum TSH levels may be increased or decreased due to NTIS or to thyroid disease (1). Serum TSH values below 0.1 mIU/L were due to hyperthyroidism in 24% with the remainder transiently reduced due to NTIS (1). TSH values below 0.01 mIU/L were due to hyperthyroidism in 73%, and to NTIS in the remainder, while only 8% with TSH values between 0.01 and 0.1 mIU/L were due to hyperthyroidism (1). Only 14% of TSH values between 6.8 and 20 mIU/L were due to primary hypothyroidism, whereas 50% of TSH values above 20 mIU/L were due to primary hypothyroidism and 50% were transiently elevated due to NTIS (1). NTIS patients with transiently elevated or reduced serum TSH levels typically had free T4 index values in the reference range for euthyroid subjects (1). Consequently, differentiating hyper- and hypothyroidism from NTIS in sick patients frequently requires assessment of a serum free T4 estimate as well as a TSH value, and may require reassessment after recovery.

ARE NTIS PATIENTS EUTHYROID OR HYPOTHYROID ?
Controversy continues as to whether patients with nonthyroidal illnesses, in the absence of concurrent hypothalamic, pituitary or thyroid disease, are hypothyroid or euthyroid (1, 2, 5-9), as summarized in Table 1. The major problems in assessing the hypothalamic-pituitary-thyroid axis in NTIS are that immunoactive TSH assays may not reflect bioactive TSH levels, and free T4 and free T3 values measured in sera from sick patients are highly method dependent (3). Systemic illnesses reduce thyroid hormone binding to serum binding proteins, causing disparities among non-dialysis free T4 estimates, which are dependent on T4 binding to serum proteins, and direct dialysis free T4 measurements, which are not (3, 10). Consequently, clinicians must be aware of the performance of the assay methods used for assessing patients with NTIS. Further, the thyroid hormone status of the tissues in patients with NTIS has not been accurately determined.

TABLE 1: THYROID HORMONE STATUS IN NTIS

Evidence supporting euthyroidism in NTIS
1) Serum TSH and free T4, by ultrafiltration of minimally diluted sera (11) or direct dialysis, are normal in the majority of NTIS patients (1, 3).
2) T4 production/degradation rates in critically ill NTIS patients may be lower than in normal subjects but are similar to those of healthy euthyroid subjects with low serum T4 binding (2). Low T4 binding in serum may result in underestimation of total T4 concentrations (12).
3) Reverse T3 production from T4 in critically ill patients with low total T4 levels are similar to those of NTIS patients with normal total T4 levels (2).
4) Free T3 concentrations in NTIS are highly method dependent and are normal in critically ill NTIS patients by ultrafiltration of minimally diluted sera (3, 11)
5) Hepatic T3 nuclear receptor proteins are increased in chronically diseased human livers (13).
6) Thyroid hormone-regulated mRNAs encoding TBG, SHBG, CBG, and transthyretin are present in normal quantities in chronically diseased human livers (14).
7) In critically ill NTIS patients, free cortisol by direct equilibrium dialysis or cortisol index values are appropriately elevated (15, 16) despite reduced total cortisol levels (4) indicating a functional hypothalamic-puitary-adrenocortical axis.

Evidence not supporting euthyroidism in NTIS
1) TRH gene expression may be reduced in NTIS (6, 17).
2) Nocturnal TSH surge and pulsatile secretion of TSH, GH and PRL are decreased, and TSH response to exogenous TRH is blunted in NTIS (1, 3).
3) T4 uptake into rapidly equilibrating tissues like liver may be reduced in critically ill patients with NTIS (2).
4) Total T3 content of cerebral cortex, hypothalamus, anterior pituitary, liver, kidney, and lung are reduced in NTIS as is hepatic content of total T4; total hormone content may reflect primarily bound hormone (18).
5) T3 production rates are decreased in all NTIS patients (2), however, total T3 levels, and, therefore, T3 production rates, may be underestimated if free T3 concentrations are normal (11).

HYPOTHALAMIC-PITUITARY FUNCTION IN NTIS
Controversy exists as to whether the fall in total T4 concentrations during acute NTIS are primarily due to central suppression indicating hypothyroidism, to reduced serum T4 binding to TBG due to protease cleavage, indicating euthyroidism, or a combination of these factors (2, 6, 19). Wadwekar and Kabadi (20) addressed this question in a study of six euthyroid men with primary hypothyroidism maintained on L-T4 replacement therapy during an acute nonthyroidal illness. Prior to the illness, total T4, total T3 and TSH levels were in the reference range for euthyroid subjects. All of these values decreased during the acute illnesses, returning to pre-illness values with recovery (20). These findings are more consistent with acutely impaired serum T4 binding than with central hypothyroidism. Since oral L-T4 replacement was continued during hospitalization, TSH suppression could not reduce thyroidal T4 production. Rather, rapid reduction of T4 binding to TBG during acute illness (19, 21) could release T4 from TBG, reduce total T4 levels, transiently increase free T4 and suppress TSH levels, as well as increasing serum T4 clearance rates and reducing serum T4 half-life values (2, 19). During recovery, normalization of T4 binding to TBG could increase total T4, and transiently reduce free T4 levels resulting in increased TSH values, until a steady state was re-established.

IS THYROID HORMONE OR SECRETOGOGUE THERAPY BENEFICIAL IN NTIS ?
Controversy continues regarding thyroid hormone therapy in NTIS (5, 7-9). Many studies utilized pharmacological rather than physiological doses of L-T4 or L-T3. Normal T4 production rates range from 80 to 100 ug per day for a 70 kg person, while normal T3 production rates range from 25 to 32 ug per day (22, 23). Absorption of L-T4 and L-T3 after oral administration varies from 50-80%. Oral L-T4 replacement doses range from 112 to 126 ug per day while oral L-T3 doses range from 42 to 56 ug per day (22, 23).
Two weeks of L-T4 therapy in physiological doses (1.5 ug/kg intravenously per day, 105 ug per day if 70 kg) given to intensive care unit patients with low total T4 concentrations did not hasten recovery or improve survival compared to placebo treated patients (7). However, pharmacological doses of L-T4 (300 ug per day intravenously) given to patients with acute renal failure for 48 hours in a prospective, randomized, double-controlled, double-blinded trial, increased mortality more than 3 fold (7).
L-T3 therapy in physiological doses (30 ug/d orally) administered during fasting did not affect protein metabolism (24, 25), while a higher physiological dose (40 ug/d orally) increased protein catabolism (26). One week of L-T3 0.8 ug/kg orally (56 ug/d if 70 kg) given to uremic patients adversely affected nitrogen balance and increased amino acid turnover (27). Pharmacological doses of L-T3 (60-150 ug/d orally) increased protein catabolism during fasting (24, 28).
L-T4 (150 ug per day intravenously) plus L-T3 (0.6 ug/kg per day intravenously, 42 ug for 70 kg) administered to critically ill NTIS resulted in TSH suppression and higher serum total T3 concentrations than in untreated patients; outcome data were not provided (29).
Pharmacological doses of L-T3 have been administered to patients with burns without affect, and after coronary artery bypass surgery or heart transplantation with improvement in cardiac function, decreases in pressor requirements, and shortened time to recovery, without an affect on overall mortality (7, 8). In an uncontrolled study, pharmacological doses of L-T3 therapy for congestive heart failure increased cardiac output and reduced systemic vascular resistance (8). Pharmacological doses of L-T3 in brain-dead heart transplant donors showed no significant differences in hemodynamic status or myocardial function (7).
TRH in combination with growth hormone secretogues given to NTIS patients with prolonged critical illnesses augmented the pulsatility of TSH release and increased total T4 and total T3 level but no clinical benefit was demonstrated (30).

CONCLUSIONS:
1) There is no compelling evidence to support overt T4 or T3 deficiency in NTIS.
2) Data do not indicate that physiological replacement doses of either L-T4 or L-T3 improve outcome in NTIS patients and such therapy may be detrimental.
3) If thyroid hormone therapy is proposed in NTIS patients without concurrent hypothalamic, pituitary or thyroid disease, large prospective, randomized, double-blinded, controlled studies should be performed in a homogeneous clinical setting with well defined and clinically relevant end-points.
4) Although pharmacological doses of L-T3 may have beneficial effects in certain circumstances, these effects are unrelated to thyroid hormone deficiency in NTIS.
5) Thyroid hormone replacement therapy in NTIS patients should be restricted to treating hypothyroidism due to documented hypothalamic, pituitary, or thyroid disease.
6) The clinician should focus on differentiating changes in serum thyroid hormone levels due to NTIS from those due to concurrent hypothyroidism or hyperthyroidism, and on treating patients with concurrent hypothyroidism or hyperthyroidism and NTIS appropriately.

REFERENCES