Evaluation of Thyroid Functions in Critically Ill

Home , Goitre

EVALUATION OF THYROID Malnutrition and illness influence every FUNCTIONS IN CRITICALLY aspect of thyroid hormone economy, from the control of secretion to the delivery, me-

ILL INFANTS tabolism and ultimate action. This has led to the terminology known as "Sick Euthyroid Syndrome" which is characterized by sig- N.K. Anand nificant decrease in serum tri-iodothyronine Vineeta Chandra (T3) slight decrease in serum thyroxin (T4) R.S.K. Sinha increase in reverse T3 level and no signifi- Harish Chellani cant change in thyroid stimulating hormone (TSH) level(l-5). During the past two de- ABSTRACT cades, considerable progress has been made The degree to which thyroid functions are in the understanding of the peripheral me- affected by non-thyroid illness and an assess- tabolism of thyroid hormones in diseases ment of its correlation with mortality was evalu- that do not primarily affect thyroid gland. ated. Thirty infants (20 M, 10 F) with a mean Interestingly, all the conditions in which age of 433±3.28 months (±1 SD), with severe sick euthyroid syndrome has been docu- acute systemic illness and 30 healthy controls, mented have nothing more in common than age and sex matched, were studied for total catabolic state. Hence, it has been suggested serum T3, T4 and TSH levels at admission and recovery or before death. Serum thyroid hor- that the decrease in thyroid hormone level mones were measured using standard techniques. may be a protective phenomenon to limit There was no significant change in thyroid protein catabolism and lower energy re- indices with age, sex, nutritional status, serum quirements in non-thyroidal illness protein and C-reactive protein. Serum T3 levels (NTI)(6). in infants were significantly lower (0.62 ± 0.63 ng/ml) than the controls (1.90 ± 0.62) The degree to which thyroid functions (p<0.001), with normal T4 and TSH levels at ad- are affected by NTI is related to the severity mission. Both serum T3 and T4 levels increased of the illness and can serve as a useful, if with recovery. Out of 30 infants studied, 14 died relatively non-specific, prognostic indica- whereas 16 were discharged. It Was noticed that tor(7). In adults, in different studies signifi- T3 and T4 values were significantly reduced at cant correlation of serum T3 and T4 levels or near death when compared with the admis- and patient's prognosis has been shown and sion levels (p<0.001). Prognosis could not be mortality is significantly higher in patients determined at the lime of admission, as thyroid indices at admission of patients who died, when of NTI with low T4 and progressively de- compared to infants who were discharged, clining T3 levels. However, in children and showed no significant difference in T3, T4 or especially in infants no definite correlation TSH levels. has yet been found(7-10). If the correlation The above mentioned changes in thyroid indices probably occur as a temporary adaptive From the Department of Pediatrics, Safdarjang mechanism to limit catabolism in states of stress Hospital, New Delhi 110 029. such as infection. Hence, it is suggested that thy- Reprint requests: Dr. N.K. Anand, Consultant roid function tests be interpreted with caution in and Head, Department of Pediatrics, patients with non-thyroid illness. Safdarjang Hospital, New Delhi 110 029. Key words: Thyroid function, Systemic illness, Received for publication: October 7, 1993; Prognosis. Accepted: May 12, 1994 ANAND ET AL. THYROID FUNCTIONS IN CRITICALLY ILL can be established, children with a poor analysis was performed. Similarly, samples prognosis (low T3 T4 levels) could be iden- were collected for the controls. Serum thy- tified earlier and this may allow for closer roxine and tri-iodothyronine were measured observation and therapeutic intervention. by radioimmunoassay technique and serum levels of thyroid stimulating hormone were Material and Methods measured by immunoradiometric assay. The The study was conducted at the Depart- radioimmunoassay kit for thyroxine (Code ment of Pediatrics in collaboration with RIAK -5/5A) and tri-iodothyronine (Code the Department of Nuclear Medicine, RIAK -4/4A) and TSH immunoradiometric Safdarjang Hospital, New Delhi. Study sub- assay kit (IRMA -9) were supplied by Board jects comprised of 30 children in age group of Radiation and Isotope Technology, 1 month to 12 months (20 M, 10 F) suffer- Bhabha Atomic Research Centre, Bombay. ing from acute severe systemic illness All the samples were tested in duplicate. requiring intensive therapy. Broncho- Protocol provided along with the kits was pneumonia was the commonest illness, strictly adhered to during analysis. others being septicemia, meningitis, gastro- enteritis, bronchiolitis and severe spasmodic Statistical analysis was done using bronchitis. All the infants were examined Student's 't' test. Null hypothesis was re- and relevant investigations and treatment jected with level of significance <0.05. for their disease were instituted and the progress monitored. Equal number of age Results (1-12 mo) and sex matched healthy infants Thirty children in the age group of 1 mo (mostly siblings of patients) were taken as to 12 mo with a mean of 4.33 ± 3.28 mo (±1 controls. All infants with maternal history SD) were studied. The outcome of patients of thyroid dysfunction, children with clini- was recorded in the form of death (n=14) or cal evidence of endocrine abnormality, discharge (n=16). The average period of especially thyroid and infants with clinical stay in the hospital in discharge and death goitre were excluded from the study. The cases was 7.6 and 2.07 days, respectively. nature and purpose of the study was care- fully explained to parents of all the subjects The mean plasma concentration of T3 and their consent was taken before includ- during sickness at admission was 0.62 ± ing them in the study. 0.63 ng/ml which was markedly low as compared to the control values (1.90 ± 0.62 Thyroid Function Tests ng/ml) (p <0.001). However, the level of Thyroxine (T4), tri-iodothyronine (T3) serum T3 came back to normal level 1.67 and thyroid stimulating hormone (TSH) ± 0.75 ng/ml at the time of discharge were measured on two occasions: (a) at the (p >0.01). Mean serum T3 levels at the time time of admission before instituting treat- of discharge when compared to levels at ment for the disease; and (b) at the time of admission (n=16), also showed a significant discharge or during the terminal stage prior difference (p <0.001). to or at death. Thyroid function indices of cases at Five ml venous blood was collected admission, death and discharge are depicted without anticoagulant and serum was stored in Table I. Serum T3 levels of patients in airtight containers at -20°C till the who died, when compared with their initial

1234

admission values, were significantly reduced. eral monodeiodination of T4 by the enzyme Also, serum T3 levels of patients at recov- 5'deiodinase. In adults and in children many ery, when compared with infants who died, medical and surgical illnesses lead to euthy- showed a significant difference (p <0.001). roid sick syndrome(l,5,7). Several studies have correlated both the serum T4 and T3 Serum T4 levels at admission were not levels with increased mortality in adults but much altered, but the levels increased sig- little is known in children especially in nificantly at the time of discharge though infants(7,10). the levels were still in the normal range. Se- rum T4 levels decreased significantly if the We observed significantly low serum T3 outcome was death. Six out of 14 infants levels in in/ants at the time of admission as who died had serum T4 values <3 µg/dl. compared to the controls (p<0.001) which The fall in serum T4 was more than fall in returned to normal at the time of discharge. T3 in those who died when compared with Twenty eight of 30 medically ill infants had the admission levels (p <0.001). Serum TSH low levels of T3 within 24 hours of admis- levels did not show any significant change sion. Serum T4 and TSH levels did not at admission and at death in patients when show any significant change at the time of compared to the controls. However, the admission. Zucker et al. (8) and Uzel and levels at discharge were higher than the Neyzi(10) had found similar changes in control values. their studies. No significant difference in T3, T4 and In children in whom the course of the TSH levels was observed with regard to sex, disease was fatal, serum T3 or T4 levels at age (<6 mo and >6 mo), nutritional status, or just prior to death were significantly low serum proteins (low and normal) and to C- as compared to their admission levels. Also reactive protein. when compared to infants who were discharged, the serum T3 values of patients Discussion who died were very low (p <0.001). This Under normal circumstances, 100% of shows that there was a marked fall in serum T4 and 10-20% of T3 in the serum are di- T3 values in infants whose illness was rectly secreted by the thyroid gland. The re- severe enough to end in mortality. Mclarty maining 80% of T3 is derived from periph- et al. in a study of 30 patients of myocardial

1235 ANAND ET AL. THYROID FUNCTIONS IN CRITICALLY ILL infarction showed a sequential and progres- related to the prognosis of NTI. If thyroid sive fall in serum T3 and T4 levels from indices are measured early in the course of the time of admission reaching abnormally critical illness, which is predictive of subse- low in all six patients who died in their quent outcome, the clinical value of such series(11). However, prognosis cannot be laboratory assessment will be enhanced be- ascertained at the time of admission. Retro- cause presumably there will be time avail- spectively, when thyroid indices at admis- able for intensive therapeutic intervention. sion of infants who died were compared to REFERENCES infants who were discharged, no significant difference could be found in serum T3, T4 1. Wartofsky I, Burman KD. Alterations in and TSH levels in the present study. In vari- thyroid functions in patients with sys- ous studies nadir value of T4 and T3 or se- temic illness: The "euthyroid sick" syn- quential decrease in the values has been cor- drome. Endocrinol Rev 1982, 3: 164-168. related to mortality. Kaptein et al. in a study 2. Feber J, Kirkeguard C, Rasmusin B, et al. of 195 critically ill medical patients corre- Pituitary thyroid axis in critical illness. J lated clinical outcome with the lowest of se- Clin Endocrinol Metab 1987, 65: 315- rial T4 values as well as other thyroid indi- 320. ces. Mortality was inversely related to nadir 3. Chopra I, Chua RTGN, Nguyen A. A serum T4 concentration(12). Our study also search of an inhibitor of thyroid hormone had 6 infants with T4 <3 µg/dl and all six of binding to serum protein in nonthyroidal them died. Uzel and Nezi in their study of illness. J Clin Endocrinol Metab 1979, 13 infants also showed serum T4 value to be 49L: 63-69. very low in 8 cases who died(10). 4. Carter JN, Eastman CJ, Corcoran JM. Ef- fect of severe chronic illness on thyroid The importance of changes in thyroid function. Lancet 1974, 2: 971-974. hormone level with non-thyroidal illness is uncertain. At present, there is no evidence 5. Gilani BB, Mac Giliway MH, Voorhess that T4 administration is required in euthy- ML. Thyroid hormone abnormalities at roid sick syndrome. Brent and Hershman diagnosis of insulin dependent diabetes mellitus in children. J Pediatr 1984, 105: studied effects of thyroxin therapy on pa- 218-222. tients with severe NTI and low serum thyroxin concentration. Thyroxin administra- 6. Richmon DA, Molitch ME, O'Donnel TF. tion rapidly normalized serum T4 concen- Altered thyroid hormones level in bacte- tration but T3 concentration did not in- rial sepsis: Role of nutritional adequacy. Metablism 1980, 29: 936-942. crease. Thyroxin therapy in the said study did not augment thyroid hormone action nor 7. Slag MF, Moley JE, Elson MK, et al. did it improve survival. Decreased conver- Hypothyroxinemia in critically ill chil- sion of T4 to T3 in the periphery has been dren. J Pediatr 1985, 107: 552-554. postulated to be the predominant cause of. 8. Zucker AA, Chernow B, Fields A. Thy- low T3 levels inspite of T4 therapy(13). roid function in critically ill children. J Pediatr 1985, 107: 552-554. Our study demonstrated prospectively that euthyroid sick syndrome occurred in 9. Zaloga GP, Cherno B, Smallridge RC. A the majority of sick infants studied and longitudinal evaluation of thyroid functions in critically ill patients. Ann Surg progressive decline in T3 and T4 values is 1985,201:456-464. 1236 INDIAN PEDIATRICS VOLUME 31-OCTOBER 1994

10. Uzel N, and Neyzi O. Thyroid functions Thyroxin metabolism in the low thyroxin in critically ill infants with infections. state of critical non-thyroidal illnesses. J Pediatr Infect Dis 1986, 5: 516-519. Clin Endocrinol Metab 1981, 53: 764- 771. 11. Mclarity DG, Ratclifte WA, Mc Coll K. Thyroid hormone levels and prognosis in 13. Brent GA- Hershman JM. Thyroxin patients with serious nonthyroidal illness. therapy in patients with severe non- Lancet 1975, 2: 275-276. thyroidal illnesses and low serum thyroxin concentration. J Clin Endocrinol 12. Kaptein EM, Grieb DA, Spencer CA. Metab 1986, 63: 1-8.

1237