Original Paper
HORMONE RESEARCH
Received: June 22, 2006 Accepted: September 21, 2006 Published online: $ $ $
Horm Res 743 DOI: 10.1159/0000XXXXX
F OO PR Changes in Autonomic Modulation to the Heart and Intracellular Catecholamines A Longitudinal Study in Differentiated Thyroid Carcinoma during Short-Term Hypothyroidism and Thyroid Hormone Replacement
L. Guasti a F. Marino a M. Casentino a M. Campanelli a E. Piantanida a L.T. Mainardi b P. Vanoli a D. De Palma a R. Bombelli a M. Ferrari a C. Crespi a C. Simoni a C. Klersy c G. Gaudio a L. Maroni a A.M. Grandi a M. Tanda a L. Bartalena a S. Cerutti b S. Lecchini a A. Venco a
© S. Karger AG, Basel
PROOF Copy for personal use only ANY DISTRIBUTION OF THIS ARTICLE WITHOUT WRITTEN CONSENT FROM S. KARGER AG, BASEL IS A VIOLATION OF THE COPYRIGHT.
a Department of Clinical Medicine, University of Insubria, Varese, b Department of Biomedical Engineering, Polytechnic University, Milan, and c Department of Biometry and Clinical Epidemiology, IRCCS S. Matteo, Pavia, Italy
Key Words Hypothyroidism Spectral analysis Baroreflex sensitivity Intracellular catecholamines
Abstract Background: The effects of thyroid deprivation on the autonomic modulation to the heart remain controversial. Methods: In this study in patients followed for thyroid carcinoma, we investigated (1) heart rate variability parameters and the baroreflex gain and (2) intracellular catecholamine levels in circulating lymphocytes during short-term hypothyroidism (phase 1) and after reinstitution of TSH-suppressive thyroid hormone replacement (phase 2). Results: The RR interval value (p ! 0.001) and systolic blood pressure (p ! 0.05) were higher in phase 1 than in phase 2. The low-frequency/highfrequency (LF/HF) ratio was significantly lower in the hypothyroid state (p ! 0.05), with a higher HF component (p ! 0.05). After adjusting for mean RR interval in the regression model, the difference between the power of RR interval oscillations calculated in the two states was greater for the LF band (p = 0.005) and it was borderline significant for the HF
© 2006 S. Karger AG, Basel 0301–0163/06/0000–0000$23.50/0 Fax +41 61 306 12 34 E-Mail
[email protected] www.karger.com
HRE743.indd 1
Accessible online at: www.karger.com/hre
band (p = 0.052). The baroreflex gain LF index was similar in the two phases. The stimulus-induced cellular production of norepinephrine and epinephrine in peripheral blood mononuclear cells was significantly higher in phase 2. Conclusion: The neurally-mediated influences on the sinus node and the study of intracellular catecholamine production suggest a reduced sympathoexcitation in hypothyroidism. The early increase in blood pressure observed after thyroid hormone withdrawal is not due to impaired sensitivity of the baroreflex arc. Copyright © 2006 S. Karger AG, Basel
Introduction
Several signs and symptoms of thyrotoxicosis, including tremors, tachycardia, an increase in blood pressure and anxiety, mimic a hyperadrenergic state, whereas clinical manifestations of hypothyroidism are suggestive of a decreased sympathetic tone [1, 2]. However, at variance with the clinical picture, experimental evidence showed that hypothyroidism can be accompanied by an Prof. Luigina Guasti Department of Clinical Medicine, University of Insubria, Ospedale di Circolo Viale Borri 57 IT–21100 Varese (Italy) Tel. +39 0332 278 111, Fax +39 0332 278 595, E-Mail
[email protected]
30.10.2006 15:41:41
Patients and Methods Patients Seventy-six consecutive patients previously submitted to total thyroidectomy and radioiodine remnant ablation for papillary thyroid carcinoma were initially enrolled. Thirty-four patients were excluded because of associated diseases (cardiac or pulmonary diseases, obesity, diabetes mellitus) and/or because on interfering pharmacological treatments (antiarrhythmic agents, antihypertensive therapy, antidiabetic agents, antidepressant drugs). Thus, 42 patients were finally included in this study and submitted to the evaluation of the baroreflex gain and heart rate variability; 24-hour urinary catecholamine excretion was measured in 27 consecutive patients, and intracellular catecholamines were studied in 15 consecutive patients. In 9 out of 42 subjects, heart rate variability parameters were not analyzed because recording of tachogram or systogram either at the first or second (see below for evaluation) was unsatisfactory. All the patients were studied twice. At first evaluation (phase 1, mean 8 SD = 18 8 4 months from the time of surgery), patients were hypothyroid following thyroid hormone withdrawal for 4 weeks to perform whole-body scan and to measure thyroglobulin. The second evaluation (phase 2) was carried out (mean 8 SD = 8 8 1.5 weeks) after reinstitution of TSH-suppressive LT4 treatment. All the patients were asked not to take tea, coffee, chocolate or cola-containing substances for at least 12 h before investigations and were studied after an overnight fasting. Moreover, they were asked to collect 24-hour urine samples several days before the two evaluations to determine urinary catecholamine levels. No patient smoked or was involved in competitive sporting activities. The study protocol was approved by the local Ethics Committee and all enrolled subjects gave their informed consent.
F OO PR
increased sympathetic neural outflow in many tissues and by different - or -adrenergic tissue responsiveness [1, 2]. In addition, clinical findings suggestive of an altered sympathetic tone in thyroid dysfunction do not seem to follow the expected changes in plasma catecholamine concentrations, since the latter are normal or decreased in thyrotoxicosis and may be normal or elevated in hypothyroidism [1–4]. Moreover, data on circulating catecholamines may differ from data on urinary catecholamine excretion, as reported for hyperthyroid patients [5]. Over the last decade, evidence has been accumulating showing the existence of an endogenous catecholaminergic system in circulating lymphocytes [6, 7]. Human lymphocytes are able to produce and store endogenous dopamine (DA), norepinephrine (NE) and epinephrine (E) in response to activating stimuli [8, 9]. In these cells, endogenous catecholamine production depends upon the activation of tyrosine hydroxylase (TH, EC 1.14.16.2), the first and rate-limiting enzyme in the synthesis of catecholamines, and is inhibited by DA through the activation of D1-like (D5) receptors [10]. It was previously suggested that lymphocyte catecholamines could represent suitable indices of long-term changes in sympathoadrenal activity [11, 12], however this issue was not investigated further until now in thyroid diseases. Studies carried out in hypothyroid patients to investigate the autonomic profile either by dosing peripheral catecholamine levels or by using spectral techniques to evaluate heart rate variability provided controversial results [1–5, 13–15]. Reported discrepancies were probably related to heterogeneity of enrolled patients in different series, showing either different degrees or length of time of hormone deprivation. In the present study we investigated patients who had short-term, severe hypothyroidism following thyroid hormone withdrawn to perform whole-body scan after total thyroidectomy for differentiated thyroid cancer. The autonomic cardiovascular regulation was assessed non-invasively using spectral analysis of heart rate variability and by computing the LF index, a spontaneous index of baroreflex sensitivity (after evaluation of the coherence between the systolic arterial pressure and the electrocardiographic RR signals) [16–25]. The aim of this study was to evaluate the spectral parameters of heart rate variability during hypothyroidism and following thyroid hormone replacement. In addition, as a secondary endpoint, we evaluated the intracellular catecholamine production in lymphocytes in the patients in the two different clinical states. 2
HRE743.indd 2
Horm Res 743
Thyroid Hormone Determination FT4, FT3 and antithyroglobulin antibodies were assayed by highly sensitive radioimmunoassay methods (Brahms GmbH, Henningsdorf, Germany), while highly sensitive immunoradiometric assays were used to measure serum thyrogloblulin (Brahms GmbH) and TSH (DiaSorin, Saluggia, VC, Italy). Normal values in our laboratory were as follows: FT4 : 0.75–1.9 ng/dl, FT3: 1.5– 5.3 pg/ml, TSH: 0.3–4.5 U/ml, thyroglobulin !1 ng/ml after total thyroidectomy, and antithyroglobulin antibodies !60 U/ml. Cell Separation and Culture Human peripheral blood mononuclear cells (PBMCs) were isolated from venous blood using heparinized tubes. Whole blood was allowed to sediment on dextran at 37 ° C for 30 min. Supernatant was recovered and PBMCs were separated by density-gradient centrifugation using Ficoll-Paque Plus. Cells were then washed twice in NaCl 0.15 M and resuspended in RPMI 1640 medium supplemented with 10% heath-inactivated fetal calf serum, 2 mM glutamine and 100 U/ml penicillin/streptomycin at a final concentration of 1 ! 106 cells/ml for subsequent culture at 37 ° C in a moist atmosphere of 5% CO2. Typical PBMC preparations contained about 80% lymphocytes and 16% monocytes. Cell viability, assessed by the trypan blue exclusion test, was always 199%. Cultured PBMCs were incubated for 48 h in the presence of phytohemoagglutinin (PHA) from Phaseolus vulgaris, 10 g/ml. Treatment duration and concentration of the pharmacological agents
Guasti et al.
30.10.2006 15:41:56
were chosen according to the results of previous experiments [8] as those which induced a maximal increase of intracellular catecholamine content. PBMCs were finally harvested and assayed for intracellular catecholamines.
F OO PR
HPLC-ED Assay of Catecholamines The catecholamines DA, NE and E were assayed by HPLC with electrochemical detection according to a previously described method [26]. Briefly, cells were centrifuged for 5 min at 1,400 g and 4 ° C, the medium was recovered and 500 l were added with an equal volume of HClO4 0.4 N, centrifuged for 5 min at 15,000 g and 20 ° C, filtered (Millex HV 13, 0.45 m, Millipore) and injected into the HPLC system for CA assay. The cell pellet was resuspended in 0.2 ml of HClO4 0.4 N and disrupted by sonication. The mixture was then centrifuged for 5 min at 15,000 g and 20 ° C and the supernatant was recovered, filtered and 30 l were injected into the HPLC system. Catecholamines in the samples were quantified by using the peak areas of a standard curve, values were normalized for medium volume or cell number, as appropriate and the values were expressed as pmol/106 cells. Urinary catecholamines were also analyzed and expressed as g/24 h. Normal values in our laboratory for urinary DA, NE and E were 72–559, 9–83, and 0–34 g/24 h respectively.
plexity of the model together with goodness of its fit to the current data. It represents a trade-off between the fit of the model (which lowers 2e) and the model’s complexity, which is measured by p. The AIC penalizes for the addition of parameters, and thus selects the model which correctly fits the data using a reduced number of parameters (i.e., a parsimonious model). Regarding both the RR series and SBP series, two frequency bands were selected for each spectrum: 0.04–0.15 and 0.15– 0.40 Hz for the low-frequency (LF) and high-frequency (HF) components, respectively [19]. The following variables were considered for subsequent evaluations: spectral powers expressed both in ms2 and normalized units (nu) and LF/HF power ratio of RR variability, spectral powers of SBP variability (mm Hg2). A cross-spectral analysis of the RR tachogram and the systogram was performed to verify the coherence (see below) in the LF band between the two signals. A non-invasive measure of the baroreflex gain (LF, ms/mm Hg) can be extracted by the ratio between power spectral densities of the RR interval and SBP series, as follows:
RR Interval and Blood Pressure Analysis All subjects were kept in a supine and comfortable position between 09:00 and 10:00 am. They were submitted to continuous electrocardiographic and blood pressure recordings (Cardioline WS2000, Remco Italia; Finapres) connected with a microcomputer, as described previously [20, 21]. Continuous Finapres systolic blood pressure (SBP) was used for subsequent analysis. Moreover, standard sphygmomanometric blood pressure was measured before the recording. We recorded a baseline period of 15 min (the last 5 min were used for analysis). As previously described, QRS detection and RR interval measurement were automatically performed by the Cardioline WS2000 equipment [20, 21]. This algorithm looked for the R wave peak as a reference point. Afterwards, each QRS complex was interpolated by a parabolic curve. The R point was chosen to correspond with the maximum of the interpolating parabola to improve the accuracy of detection of the peak R wave [16]. The SBP value was automatically identified as the maximum of the parabolic curve fitting the pressure tracing. Corrections made on RR intervals determined automatic corrections in the blood pressure series. In this way, a series of successive RR intervals (RR tachogram) and a series of corresponding successive SBP values (systogram) were obtained. The power spectrum analysis was performed by an autoregressive algorithm on a 3–5 min recording, after a visual identification of the tachogram. The order was automatically selected (AIC criterium) [16, 17] in a range between 8 and 15, 8 being the most frequently used. The Akaike Information Criterion (AIC) has become a standard tool in time series model fitting. It may be used to estimate the ‘optimal’ number of parameters (i.e. the socalled ‘model order’) of an AR model. The preferred model order p is that which minimizes the following expression:
BLF =
PRR -LF
PSBP -LF
where PRR-LF is the power (ms2) in the LF band obtained by the tachogram and PSBP-LF is the power (mm Hg2) in the LF band obtained by the systogram. The amount of linear coupling between two signals in the frequency domain can be expressed by means of the coherence function. We used the k 2 value, i.e. the squared coherence values. For all the considered spectra the coherence in the LF band was 1 0.5. Statistical Analysis Descriptive statistics were computed as mean and standard deviation for continuous variables. For skewed distribution, median and interquartile range (IQR) were used. A regression model for repeated measures was used to compare RR and blood pressure variability at baseline and after treatment with LT4. Robust standard errors were computed to account for intra-patient correlation of measures. Effect estimates were adjusted for intracellular catecholamines (baseline and after stimulus with PHA), and for mean RR value in bivariate analysis. For the purposes of the analysis, log transformation was applied for the following variables: LF power, HF power and LF/HF ratio for both RR and blood pressure variabilities, LF, catecholamines, TSH, FT3, FT4. Stata8 (Stata Corp., College Station, Tex., USA) was used for computation. A two-sided p value ! 0.05 was considered statistically significant.
Results
where 2e is the variance of the prediction error and N is the number of samples. The idea behind the AIC is to examine the com-
At first evaluation, serum FT3, FT4 and TSH concentrations were 1.35 (IQR: 1.0–1.6) pg/ml, 0.20 (IQR: 0.1– 0.3) ng/dl, and 87.4 (IQR: 56.0–100.0) U/ml, respectively, whereas values at second evaluation (when TSH-suppressed thyroid hormone therapy) were as follows: FT3
Hypothyroidism and Spectral Analysis
Horm Res 743
AIC (p) = N ln(2e (p)) +2 p
HRE743.indd 3
3
30.10.2006 15:41:56
Table 1. RR and SBP variability in hypothyroidism and after thyroid hormone replacement
Hypothyroid state median (IQR)
LT4 replacement median (IQR)
p*
LFRR, ms2 HFRR, ms2 LFRR, nu HFRR, nu RR LF/HF LFBP, mm Hg2 HFBP, mm Hg2
122.1 (55.2–179.6) 94.3 (34.4–209.6) 54.7 (27.3–45.8) 39.2 (26.5–58.6) 1.4 (0.4–2.5) 3.5 (1.6–5.5) 1.9 (1.4–2.9)
157.9 (101.9–285.1) 84.8 (37.8–162.2) 61.6 (43.8–74.0) 31.4 (15.8–46.6) 1.9 (1.0–4.7) 2.9 (1.5–6.7) 1.9 (1.1–3.0)
0.08 0.46 0.06 0.04 0.05 0.75 0.45
F OO PR
Variable
* p values are calculated for log-transformed variables.
Table 2. Intracellular catecholamines
and urinary catecholamines during the hypothyroidism state and after thyroid hormone replacement
Variable
Hypothyroid state LT4 replacement median (IQR) median (IQR)
p*
Dopamine 0 h, pmol ! 106 cells Dopamine 48 h PHA, pmol ! 106 cells Norepinephrine 0 h, pmol ! 106 cells Norepinephrine 48 h PHA, pmol ! 106 cells Epinephrine 0 h, pmol ! 106 cells Epinephrine 48 h PHA, pmol ! 106 cells Urinary dopamine, g/24 h Urinary norepinephrine, g/24 h Urinary epinephrine, g/24 h
0.64 (0.16–1.06) 1.61 (0.10–9.10) 0.57 (0.14–1.34)
0.41 0.2 0.87
0.38 (0.18–0.48) 6.78 (5.94–9.58) 1.06 (0.16–1.48)
8.56 (0.16–13.02) 15.22 (5.24–18.64)
