The Prognostic Value of Severe Malnutrition in the Development of Nonthyroidal Illness in Head and Neck Cancer Patients

Nonthyroidal illness (NTI), also known as the euthyroid sick syndrome, is characterized by increased serum concentrations of the metabolically inactive hormone reverse T^sub 3^ (rT^sub 3^) and a low serum concentration of 3,3',5-triiodothyronine (T^sub 3^). Besides these changes, total serum concentrations of thyroxine (T^sub 4^) can be reduced in severely ill patients. The free T^sub 4^ (FT^sub 4^) concentration, however, may be reduced or increased. Thyrotropin (TSH) concentrations are usually normal, but in 10%-15% of patients with low concentrations of T^sub 4^, TSH plasma levels are decreased.1 The degree of hormonal disturbance correlates with the severity of the underlying condition.2 Considering the serum T^sub 3^ concentration as one of the major factors in determining the basal metabolic rate, the development of NTI can be regarded as an adaptive response of the human body to save energy by conserving proteins and protecting organ function.3 Otherwise, patients with NTI having low serum concentrations of T^sub 4^ have a high probability of death, indicating that severe NTI may be harmful.4 In fact, when serum concentrations of T^sub 4^ drop below 4

NTI has multiple causes, including surgery, trauma, inflammatory conditions, myocardial infarction, severe illnesses like liver disease, chronic renal failure and malignancy, administration of some drugs, and also starvation.4,9,10 Starvation decreases T^sub 3^ concentrations, whereas short-term overfeeding increases plasma T^sub 3^ levels, demonstrating the modulation of thyroid metabolism by caloric content.11-13 Interestingly, thyroid hormone metabolism is also influenced by dietary composition.14 Carbohydrate restriction mimics the effects of starvation on T^sub 3^ and rT^sub 3^ concentrations,11,15 and refeeding with diets containing carbohydrates but no fat reverses the hormonal changes induced by starvation.16,17 Because all mentioned studies are executed in healthy individuals, little is known about the combined effect of reduced food intake and disease on thyroid function. Therefore, we investigated whether the degree of NTI due to surgery differed between severely malnourished and well-fed head and neck cancer patients.

MATERIALS AND METHODS

The study was approved by the institutional review boards, patients were informed of the purpose of the study, and written informed consent was obtained from all participating patients.

Patients

Between January 1, 1994, and December 31, 1997, 22 severely malnourished patients (weight loss >10% during 6 months before surgery18), admitted to the Department of Otolaryngology/Head and Neck Surgery of the VU University Medical Center in Amsterdam, The Netherlands, for intentional curative surgery of T1-T4 carcinomas of the head and neck, were prospectively followed. All patients had a histologically proven squamous cell carcinoma of the oral cavity, larynx, oropharynx, or hypopharynx and underwent major head and neck surgery. Patients were not allowed to participate in other trials at the same time. Other exclusion criteria were thyroid dysfunction in the year before surgery and hemi- or total thyroidectomy during the operation. Twenty-two patients entered the study. Five patients underwent in addition a hemithyroidectomy in combination with a total laryngectomy, and 1 patient had severe thyroid hormonal disturbances before surgery. Therefore, these patients were excluded. After exclusion of these patients, 16 malnourished patients were studied. Six well-fed head and neck cancer patients eligible for surgical treatment served as a control group. The surgical procedure in this group was comparable with the operation of the malnourished group. All patients were postoperatively fed with enterai nutrition (Nutrison, Nutricia Nederland B.V., Zoetermeer, The Netherlands) which consisted of 62.50 g protein, 48.61 g fat, and 140.63 g carbohydrate per L. Nutrition support contained approximately 150% basal energy expenditure, as calculated with the Harris-Benedict formula.19

Laboratory Tests

Plasma levels of the thyroid hormones (rT^sub 3^, T^sub 3^, FT^sub 4^, and TSH) were measured on the day before the operation and on the first, fourth, and seventh day after the operation. rT^sub 3^ was measured by radioimmunoassay (CLB, Academic Medical Center, Amsterdam, The Netherlands). T^sub 3^ and FT^sub 4^ were measured by competitive immunoassay (ACS: 180 system, Chiron Diagnostics, Emeryville, CA). TSH was measured by immunometric assay (ACS: 180 system, Chiron Diagnostics, Emeryville, CA).

Statistical Analysis

Mann-Whitney U test was used to investigate differences among various time points between the 2 groups, and Wilcoxon signed-rank test was used to investigate within-group differences between the time points. For each comparison, the overall a level was set at .05. A Bonferroni post hoc a-level adjustment for multiple comparisons was made for both tests. Data are presented as means ± SEM. Statistical analysis was performed using SPSS 11.0 for Windows (SPSS Inc. Chicago, IL).

RESULTS

Patient characteristics and their oncologic status are presented in Tables I and II, respectively. There were no significant differences in men:women ratio, length, albumin (4 time points), tumor stage, tumor localization, and comorbidity in the 2 groups. The malnourished patient group had a significantly lower age, weight, and body mass index compared with the wellfed-patient group. In comparison with the well-fed patients, relatively more tumors of the oral cavity and oropharyngeal area occurred in the malnourished patient group, which could possibly be an explanation for their malnutrition. In addition, mean duration of the operation and mean recorded blood loss during the operation did not differ between groups. The mean postoperative amount of fluid infusion was also not different between both groups.

Perioperative Thyroid Hormone Changes

The course of rT^sub 3^, T^sub 3^, and FT^sub 4^ is shown in Figures 1, 2, and 3, respectively. The baseline concentration of the hormones rT^sub 3^, T^sub 3^, and FT^sub 4^ did not significantly differ between the 2 groups. rT^sub 3^ increased significantly (p = .009) in malnourished patients, whereas there was no significant increase in well-fed patients from the day before the operation to the first postoperative day. In addition, T^sub 3^ decreased significantly (p = .001) in the malnourished patient group from day - 1 till day 1 but showed no statistically significant reduction in well-fed patients. In malnourished patients, FT^sub 4^ declined significantly (p = .001) after the operation, whereas no significant reduction of FT^sub 4^ was seen in well-fed patients. TSH showed no significant change either in malnourished or in well-fed patients.

Postoperative Hormone Changes

From day 1 till day 4 of the postoperative period, rT^sub 3^ showed no significant change in malnourished patients. Interestingly, rT^sub 3^ was still significantly (p = .002) higher on day 4 compared with the day before the operation. On day 7, no significant difference compared with the day before the operation was seen. In well-fed patients, rT^sub 3^ showed no significant changes during the whole postoperative period.

In the malnourished group, both T^sub 3^ and FT^sub 4^ showed a significant increase between days 1 and 4 (p = .01 and p = .002, respectively), but no significant changes of both hormones were seen between day 4 and day 7.

In well-fed patients, no significant changes of both T^sub 3^ and FT^sub 4^ were seen during the whole postoperative period.

DISCUSSION

This study shows that head and neck surgery is accompanied by a significant increase of rT^sub 3^ and a significant decrease of T^sub 3^ and FT^sub 4^ in malnourished patients, indicating the development of NTI in these patients. In addition, on day 4, rT^sub 3^ was still significantly higher compared with the day before the operation, but on day 7 no significant difference with the day before the operation was seen. Furthermore, it took 4 days till the hormones T^sub 3^ and FT^sub 4^ were restored in malnourished patients. In contrast, none of the well-fed patients developed NTI.

The observed changes in thyroid hormone concentrations and the period in which they restore are described in many other studies. However, these studies never distinguish between well-fed and severely malnourished patients undergoing major head and neck surgery. Interestingly, Richmand et al20 studied the effect of nutrition support on thyroid hormone changes in septic patients. They concluded that the alterations in thyroid hormone levels found in these patients are largely caused by caloric deprivation (600-1000 kcal/d of 5% dextrose in water) associated with such severe illness. Malnourished patients have a low protein state, which may affect serum binding of thyroid hormones to thyroxine binding globulin (TBG) and thyroxine binding prealbumin. This may result in a decrease in the total level of thyroid hormones, without changing the free fractions. However, in our malnourished patient group not only did the total thyroid hormone levels change, but FT^sub 4^ also decreased significantly in the malnourished patient group after the operation. Therefore, reduced protein binding of thyroid hormones in the malnourished patients does not seem to be a logical explanation for the difference in the development of NTI between malnourished and well-fed patients in our study.

Interestingly, serum cortisol levels increase during starvation, and it is known that the TBG concentration is decreased by high glucocorticoid levels. In addition, cortisol reduces activity of type I iodothyronine deiodinase, thereby inhibiting the generation of T^sub 3^ and preventing the metabolism of rT^sub 3^ in the liver. This enzyme is also affected by other factors such as cytokines and unsaturated fatty acids.21-23 Interleukin-6 (IL-6) is a ubiquitous cytokine and is secreted under many circumstances like surgery, inflammation, and exercise.21 It has many endocrine functions, including activation of corticotropin and cortisol release.24 Torpy et al22 studied the effect of IL-6 on the thyroid axis in healthy humans. After 24 hours, they found reduced T^sub 3^ and elevated rT^sub 3^ levels. In addition, cortisol levels were greatly elevated 2 hours after IL-6 injection compared with control values. They concluded that the alterations in thyroid hormone metabolism caused by IL-6 may be due to a direct inhibition of type I iodothyronine deiodinase or an indirect effect of elevated cortisol levels on this enzyme.22 Because we did not measure cytokine profiles or cortisol levels, no conclusion can be drawn about their influence on thyroid hormone metabolism in our study.

Another factor that could be responsible for our findings could be changed fat metabolism in malnourished patients. Chopra et al21 showed a significant correlation between serum concentrations of free unsaturated fatty acids and the inhibition activity of type I iodothyronine deiodinase in patients with NTI. The mechanism by which fatty acids inhibit type I iodothyronine deiodinase is not clearly understood, but a possible explanation is alteration of the hydrophobic environment of the enzyme.25

Another possible explanation for differences in the development of NTI between malnourished and wellfed patients is the reduction of tissue energy. It has been demonstrated that thyroid hormone transport into cells requires energy.26 During malnutrition ATP concentrations may be low, resulting in a lower T^sub 4^ transmembranal transport into T^sub 3^ producing tissues. As known, the liver plays a major role in the production of T^sub 3^.1 Thus, inhibition of T^sub 4^ into the liver will reduce plasma T^sub 3^ levels, de Jong et al26 studied the effect of IV administration of fructose, which reduces intracellular ATP, in humans. This induced a rise in serum lactic and uric acid, indicating a decrease in ATP in the liver. In addition, they observed a concomitant decrease in serum T^sub 4^ tracer disappearance during the first 3 hours after administration of fructose. The authors concluded that the uptake of T^sub 4^ into the liver might be regulated by intracellular energy supplies.

Because it is known that the degree of thyroid hormonal disturbance of patients correlates with mortality, it might be important to prevent these hormonal disturbances. A randomized prospective study was done by Brent and Hershman. They administered T^sub 4^ to 12 patients with severe NTI. T^sub 4^ concentrations significantly increased but did not normalize T^sub 3^. Importantly, mortality rates did not differ between both groups. In addition, Becker et al28 studied severely burned patients who were treated with 200 µg T^sub 3^, and results showed neither a beneficial nor a disadvantageous effect. Furthermore, different studies with animals that were treated with T^sub 3^ or T^sub 4^ were not able to prove a beneficial or disadvantageous effect.29-31

In conclusion, our study shows that peri- and postoperative changes of rT^sub 3^, T^sub 3^, and FT^sub 4^ change significantly in malnourished patients compared with wellfed patients. Therefore, it can be concluded that malnourished patients are prone to develop severe NTI after major head and neck surgery. The results of our study, in combination with the insufficient evidence that replacement therapy with T^sub 3^ and T^sub 4^ is favorable for treating NTI, suggest that the preoperative nutrition status of patients who will undergo major head and neck surgery should be optimized in order to prevent severe postoperative thyroid hormonal disturbances.