| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrine Care |
Cattedra & Sezione di Endocrinologia, Dipartimento Clinico-sperimentale di Medicina e Farmacologia, University of Messina School of Medicine, 98125 Messina, Italy
Address all correspondence and requests for reprints to: Salvatore Benvenga, M.D., Cattedra & Divisione di Endocrinologia, Policlinico Universitario di Messina, Padiglione H, 4 piano, 98125 Messina, Italy. E-mail: s.benvenga{at}me.nettuno.it
Abstract
Old studies in animals and unblinded studies in a few hyperthyroid patients suggested that L -carnitine is a periferal antagonist of thyroid hormone action at least in some tissues. This conclusion was substantiated by our recent observation that carnitine inhibits thyroid hormone entry into the nucleus of hepatocytes, neurons, and fibroblasts. In the randomized, double-blind, placebo-controlled 6-month trial reported here, we assessed whether 2 or 4 g/d oral L-carnitine were able to both reverse and prevent/minimize nine hyperthyroidism- related symptoms. We also evaluated changes on nine thyroid hormone-sensitive biochemical parameters and on vertebral and hip mineral density (bone mineral density). Fifty women under a fixed TSH-suppressive dose of L -T4 for all 6 months were randomly allocated to five groups of 10 subjects each. Group 0 associated placebo for 6 months; groups A2 and A4 started associating placebo (first bimester), substituted placebo with 2 or 4 g/d carnitine (second bimester), and then returned to the association with placebo. Groups B2 and B4 started associating 2 and 4 g/d carnitine for the first two bimesters, and then substituted carnitine with placebo (third bimester).
Symptoms and biochemical parameters worsened in group 0. In group A, symptoms and biochemical parameters worsened during the first bimester, returned to baseline or increased minimally during the second bimester (except osteocalcin and urinary OH-proline), and worsened again in the third bimester. In group B, symptoms and biochemical parameters (except osteocalcin and urinary OH-proline) did not worsen or even improved over the first 4 months; they tended to worsen in the third bimester. In both the A and B groups, the two doses of carnitine were similarly effective. At the end of the trial, bone mineral density tended to increase in groups B and A (B > A).
In conclusion, L-carnitine is effective in both reversing and preventing symptoms of hyperthyroidism and has a benefical effect on bone mineralization. Because hyperthyroidism depletes the body deposits of carnitine and since carnitine has no toxicity, teratogenicity, contraindications and interactions with drugs, carnitine can be of clinical use.
L-CARNITINE IS A quaternary amine
(ß-hydroxy-
- trimethylammonium butyrate) that is ubiquitous in
biological fluids and tissues of mammals, where it plays an important
role in energy metabolism (1, 2, 3, 4). Primary and secondary
deficiencies of carnitine, including the depletion of cardiac carnitine
associated to coronary heart disease and heart failure, are the
therapeutic indications for carnitine (1). Oral doses
usually range between 1 and 4 g per day (1).
Old studies in animals showed that carnitine is capable of contrasting thyroid hormone-driven changes associated with the metamorphosis of tadpoles and the nitrogen balance of rats (5, 6, 7). The same group of authors (7) also showed that serum and liver concentrations of alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were increased in rats treated with T4, but were decreased in rats treated with carnitine. These studies were followed by unblinded studies on a small number of thyrotoxic patients, who were treated solely with 1–3 g per day oral carnitine for a few weeks (8, 9, 10, 11). Although quantification of the symptomatology and statistical analysis were lacking, the authors (8, 9, 10, 11) reported a definitive improvement in the symptomatology starting from the second week of treatment; goiter size, thyroid [131]I uptake, ophthalmopathy, and serum protein-bound iodine were unchanged (8, 9, 10, 11). Carnitine was, thus, considered to be a peripheral antagonist of thyroid hormone action, not an inhibitor of the thyroid gland function (8, 9, 10, 11).
Considering that until now we lack an ideal antagonist of thyroid hormone action, it is quite surprising that no subsequent studies were performed. We have developed an interest for carnitine and have found that it inhibits thyroid hormone entry into the nucleus of human and animal cells (fibroblasts, hepatocytes, neurons) (12), thus explaining the peripheral antagonism. Based on this experimental evidence (12), we wished to conduct a controlled trial to test the clinical use of carnitine.
Patients and Methods
Generalities
We tested the possible benefits of carnitine in adult women (namely, the vast majority of thyroid patients) under TSH-suppressive doses of L-thyroxine for benign nodular goiter. We preferred these patients to Graves’ disease patients for several reasons, among which the additional drugs and/or nonmedical modalities of therapy that the latter patients might require, thus complicating the study. Moreover, with our protocol we could evaluate a double effect of carnitine, namely whether it was capable of both reversing and preventing hyperthyroidism.
In pilot studies on humans, we observed that carnitine does not antagonize the physiological negative feedback of thyroid hormones on TSH secretion.
Study design (protocol)
This randomized, double-blind, placebo-controlled trial with cross-over between placebo and carnitine had been approved by the Ethical Committee of our University. Criteria for entry, after signature of the informed consent form, were: 1) to be in good health, based on thorough physical examination and routine clinical chemistry; 2) no previous use of thyroid hormones and carnitine; and 3) no use of other medications.
Patients were randomly allocated to three major groups of 10 persons each: 0 (zero, meaning no carnitine at any time), A and B. All patients, who were euthyroid at baseline, received 2.0–2.4 µg per kg body weight per day (13) oral L-thyroxine (Eutirox; Bracco, Milan, Italy), which was taken 2–3 h before breakfast to maximize intestinal absorption (14). Although not necessary in a cross-over study, group 0 served as a surrogate group, namely it served to show what could have happened to group A and B patients if treated with L-thyroxine alone. Groups A and B consisted of two subgroups each (A2, A4, B2, B4), based on the daily oral dose (2 or 4 g, when indicated) of L-carnitine; clearly, the two subgroups served to evaluate possible dose-dependency of the carnitine effects. L-carnitine (Carnitene, enteral liquid vials of 1 or 2 g; Sigma-Tau, Pomezia, Italy) was taken twice daily (1 vial after lunch, 1 vial after dinner). In group A, L-thyroxine was associated with placebo from d 1 through 60 or with carnitine during d 61 through 120 or again with placebo during d 121 through 180. In group B, L-thyroxine was associated to carnitine during d 1 through 120 or placebo during d 121–180. We decided to administer carnitine for 4 months in group B to have a consistent evidence of its effectiveness (4 months, not 2 months).
At day 0, 30, 60, 90, 120, 150, and 180 (± 1 day) all 50 patients were
evaluated clinically and biochemically (see below). In addition to
tablet counting, compliance to L-thyroxine was assessed by
measuring serum free thyroid hormones (FT3,
FT4) and TSH (which had to be 
0.01 mU/l) at all
visits. The three hormones were measured with the
electrochemiluminescent assay by Boerhinger (Mannheim, Germany). The
corresponding coefficients of variation (CV) are 2.8%, 2.5%, and
2.2% (intra-assay) and 3.9%, 3.8%, and 3.3% (interassay). In
addition to vial counting, compliance to carnitine was assessed by
measuring the 24-h urinary excretion of carnitine (3)
(courtesy of Drs. A. Toscano and M. d’Agennouz, Clinica Neurologica 2,
University of Messina), with patients being not aware of this
particular scope of the urine collection. Figure 1
shows the compliance to the two drugs
and, at the same time, the homogeneity of the five groups in terms of
serum hormones.
|
Evaluation of the symptomatology
Other authors (cf. 15) have discussed the unnecessary complications of certain "clinical indices" in favor of simple clinical rules. Clinical indices are numerous (e.g. Refs. 16, 17, 18, 19), as none is satisfactory. The prototype of such indices is the Crooks’ index (20), which is a single number resulting from a scoring system based on the presence or absence of some symptoms and signs. Present symptoms are assigned different scores (e.g. dyspnea +1, palpitations +2, weight decrease +3), which are fixed, so that changes in their severity cannot be quantified.
Based on our experience on hundreds of naturally hyperthyroid patients
under antithyroid drug therapy and thousands of patients under TSH
suppressive-therapy, we have constructed the 5-point scale presented in
Table 1 to quantify symptoms. Compared
with the Crooks’ index (20), we have omitted sweating,
because enrollment and follow-up of different patients occurred in
different seasons, and have added insomnia and knee reflexes. On 100
overtly hyperthyroid adult women whose FT4 had
been measured with the same assay used in the present study, the
correlation of our scoring system with serum FT4
was 0.41 (P < 0.001, dyspnea) to 0.78
(P < 0.001, palpitations) (unpublished data),
which are r values better than the 0.36 (P level
not given) of the Crooks’ index (20) and comparable with
the 0.57–0.70 values (P level not given) of the Geffner’s
index (17). In these 100 patients, the score for each
subjective symptom decreased, on the average, by 3-fold or more
(P < 0.02 to P < 0.001 by two-tailed
paired Student’s t test) once euthyroidism has been
restored by antithyroid drug therapy (unpublished data). The
magnitude of changes is similar to the variation in 10 symptoms (scored
on a 5-point scale from 0 to 4) reported by Klein et al.
(16) for their 10 medically treated hyperthyroid patients.
Symptoms worsened again in the 36 of 100 patients whose hyperthyroidism
relapsed.
|
Biochemical evaluation
Although oral doses up to 15 mg L-carnitine per day
are well tolerated (1), nevertheless we monitored routine
blood chemistry and urine analysis monthly. Among the blood parameters,
some are regulated by thyroid hormones. ALT, AST,
-glutamiltransferase (GGT), and alkaline phosphatase are
up-regulated, whereas creatine phosphokinase and cholesterol are
down-regulated (21). All these parameters were measured
with the IL 900 Autoanalyzer (Instrumental Laboratory, Milan, Italy)
using colorimetric kits having intra-assay CV less than 4% and
interassay CV less than 6%. For the purpose of this study, sera were
stored at -20 C until the trial was completed. Then, each parameter in
the 350 samples (7 samples/patient x 10 patients/group x 5
groups) was measured in the same run.
Also, after the end of the trial and in a single run, four additional parameters, all up-regulated by thyroid hormones, were measured: sex hormone-binding globulin (SHBG), ferritin, osteocalcin, and the 24-h urinary excretion of hydroxyproline (OH-P) (21). Serum SHBG and osteocalcin were measured with the imunoradiometric assay by Radim (Pomezia, Italy) and CIS (Gif-sur-Yvette, France), respectively; ferritin with the chemiluminescent assay by Diagnostic Products (Los Angeles, CA); and OH-P with the high-performance liquid chromatography kit by Bio-Rad Laboratories, Inc. (Hercules, CA). The corresponding intra-assay CV were 5.1%, 3.8%, 5.2%, and 5.2%, whereas the interassay CV were 5.2%, 4.7%, 8.2%, and 6.6%.
Bone density evaluation
Thyroid hormones have a dual effect on bone: stimulation of osteoblasts, with the subsequent increase of markers of bone formation such as serum osteocalcin, and stimulation of osteoclasts, with the subsequent increase of markers of bone resorption such as urinary OH-P (21). Because of this and because of the controversy surrounding the relationship between thyroid hormone treatment on bone (22), we thought it was of interest to complement the biochemical measurements of serum osteocalcin and urinary OH-P with a clinically relevant parameter: bone mineral density (BMD). BMD (g/cm2) of the lumbar spine (L2–L4) and left femur was measured at baseline and end of the trial by dual-energy x-ray absorptiometry using a Hologic QDR-2000 instrument (Waltham, MA). The second or posttreatment BMD could not be always scheduled at the 180th day (±1 day) (viz. the same day of the last visit and blood and urine collection). For this reason, two to four patients in each group did not report themselves to the second BMD. Consequently, the two BMD measurements are available for six to eight subjects per group.
Adverse events
Undesirable clinical symptoms reported by patients were recorded at each visit starting from d 1, based on direct clinical questioning. In addition to vital sign assessment, physical examinations and blood and urine chemistry were conducted at all visits.
Statistical analysis
Data are presented as mean ± SE; the 95% confidence interval (CI) for changes is also given. Differences between means were analyzed by ANOVA.
The level of statistical significance was set at P less than 0.05.
Results
Changes in serum FT3,
FT4, TSH, and 24-h urinary excretion of carnitine
at entry and during the trial are reported in Fig. 1
. As expected, all
50 patients were confirmed to be euthyroid at baseline by thyroid
function testing. Differences between groups concerning
FT3, FT4, and TSH being not
statistically significant (P > 0.05). Serum
FT3, FT4, and TSH at entry
ranged 4.1 ± 0.2 (group B2) to 4.5 ± 0.2 pmol/l (group 0),
14.8 ± 0.7 (group A2) to 16.5 ± 0.1 pmol/l (group A4), and
1.4 ± 0.4 (group B2) to 1.9 ± 0.5 mU/l (group A2),
respectively. At the 1-month visit and thereafter, in all five groups
serum FT3 was in the upper-normal range, serum
FT4 was above the upper normal limit, and serum
TSH was consistently suppressed (Fig. 1). Again, there were no
intergroup statistical differences in the three hormone levels
(P > 0.05) throughout the trial. Urinary excretion of
carnitine at entry ranged from 206 ± 37 (group 0) to 251 ±
4.1 µmol/24 h (group B4) (P > 0.05). Urinary
carnitine peaked at the time expected (viz. months 3 and 4
in both groups A2 and A4; months 1 through 4 in both groups B2 and B4),
and with the gradient expected (A4 > A2, and B4 > B2) (Fig. 1), proving that patients were compliant to carnitine
administration.
Reversal effect of carnitine
Symptomatology. Modifications of the clinical parameters in
groups 0, A2, and A4 are shown in the top three panels of
Figs. 2-4.
The first eight parameters are positively regulated by thyroid
hormones; thus, increase indicates worsening. Body weight is negatively
regulated; thus, worsening is indicated by a decrease in kilograms.
Over the first 2 months of the trial symptoms/signs worsened to a
similar extent in the three groups. However, the profile of changes
diverged during the following 2 months in group 0 (symptoms increased)
vs. groups A (symptoms decreased). When placebo was
reintroduced in place of carnitine (fifth and sixth month of therapy),
symptoms worsened (or tended to) in groups A2 and A4. Overall, during
the 6 months of the trial, the intensity of each symptom/sign in the A2
and A4 groups was statistically different from group 0
(P < 0.05 to P < 0.001) (
Figs. 2–4);
differences concerning body weight were not significant
(P > 0.05).
|
|
|
, the important variations are
presented as percentage changes on the same scale for all nine clinical
parameters. In these figures, changes brought about by carnitine during
the third and fourth month with respect to the first and second month
(i.e. the first placebo period) in patients of groups A2 and
A4 are contrasted with the equivalent changes in patients of group 0
(who continued to take placebo, instead of switching to carnitine).
Differences between the two carnitine groups and the placebo group were
statistically significant (P < 0.01 to
P < 0.001), but differences between the two carnitine
groups were not (P > 0.05), except for asthenia
(P < 0.05). The parameters that changed most were
palpitations [35% (CI -48.6 to -21.4%)] and nervousness [33%
(CI -49.9 to -15.5%)] in group A2, and asthenia [43% (CI -48.3
to -37.3%)], palpitations [39% (CI -44.7 to -32.8%)] and
nervousness [31% (CI -41.5 to -20.1%)] in group A4.
|
Biochemical parameters. With the exception of TSH (which was
illustrated in Fig. 1), creatine phosphokinase and cholesterol, the
other biochemical parameters are up-regulated by thyroid hormones. The
peripheral parameters in groups 0, A2, and A4 are illustrated in the
top three panels of Figs. 6-8. Taking
into account all data from 0–6 months for any given parameter, groups
A2 and A4 were statistically different from group 0 (P
< 0.05 to P < 0.001) (
Figs. 6–8). Over the first 2
months, the positively regulated parameters increased in the three
groups. However, similarly to the clinical parameters (see above), the
profile diverged during the subsequent 2 months in group 0
vs. groups A2 and A4, except for cholesterol, osteocalcin,
and urinary OH-P.
|
|
|
, and they were statistically
significant (P < 0.05 to P < 0.001)
but with two exceptions. In fact, variations of cholesterol overlapped
(P > 0.05) in groups 0 [1.0% (CI -1.7 to 3.7)], A2
[4.0% (CI 0.4–7.5)] and A4 [3.7% (CI 0.43–6.9)], as did
variations in OH-P, a marker of bone resorption [3.9% (CI 2.2–5.5),
5.1% (CI 1.8–8.3), and 4.4% (CI 2.2–6.5%), respectively] (Fig. 9). In contrast, the positively regulated osteocalcin, which is a
marker of bone formation, increased to a greater extent
(P < 0.001) in groups A2 and A4 [17.6% and 19.5%
(CI 13.9–21.3 and 15.1–23.9)] compared with group 0 [7.0% (CI
4.4–9.6)]. For all parameters, differences between the two doses of
carnitine were not statistically significant (P >
0.05).
|
Bone density. Because of the small number of subjects, the
increase of posttherapy BMD of either lumbar spine or left proximal
femur in groups A was not statistically significant (P
> 0.05) from the corresponding posttherapy changes observed in group 0
vs. groups A (Fig. 10).
|
Symptomatology. Modifications of the clinical parameters
throughout the clinical trial in groups B2 and B4 are illustrated in
the bottom two panels of
Figs. 2–4, and they can be
contrasted with the corresponding modifications in group 0 (top
panel of
Figs. 2–4). Except for body weight, the 6-month profile
of each parameter (i.e. overall increase) in the placebo
group was statistically different (P < 0.05 to
P < 0.001) from the two carnitine groups. In these two
groups, symptomatology overtly decreased as long as carnitine was
maintained, and there was a tendency to worsening during months 5 and
6, when carnitine was replaced by placebo (
Figs. 2–4).
In Fig. 11 percentage changes over
baseline of the severity of symptoms/signs during the 4 months of
adjunctive carnitine treatment in groups B2 and B4 are contrasted
with the corresponding changes in group 0 patients, in whom the
adjunctive treatment was placebo. Differences between the B2 and B4
groups vs. the 0 group were significant (P
< 0.05 to P < 0.001). The largest changes concerned
nervousness [26% (-39.6 to -12.8%)] and asthenia [22% (-36.2
to -8.4%)] in group B2, and asthenia [32% (-40.5 to -23.6%)],
nervousness [24% (- 40.6 to -6.9%)], palpitations [23% (-37.0
to -8.4%)] and insomnia [-23% (-37.0 to -9.2%)] in group B4.
Heart rate decreased by 2.9% (CI -5.2 to -0.6%) or 3.6% (CI -6.1
to -1.2%) in group B2 or B4, respectively. Similarly to the A groups,
the two doses of carnitine were equally effective (Fig. 11), and
amelioration lagged 1 or 2 wk behind commencement of carnitine.
|
, whereas
the relevant percent changes are summarized in Fig. 12. Except for TSH (see Fig. 1) and
urinary OH-P, absolute values of the other parameters throughout the 6
months of the trial (
Figs. 6–8) and percentage changes during the
first 4 months with respect to baseline (Fig. 12) were in group 0
statistically different from groups B2 and B4 (P <
0.05 to P < 0.001). Confirming the observation in
groups A2 and A4 vs. group 0, serum osteocalcin increased
over baseline to a greater extent (P < 0.001) in
groups B2 and B4 vs. group 0, namely by 32% (CI 25–38.2)
and 36% (CI 27–44.6) vs. 22% (CI 19.3–24.9) (Fig. 12).
Except for ferritin (P < 0.01) and SHBG
(P < 0.05), the two doses of carnitine were equally
effective (Fig. 12).
|
). However, when either carnitine group
is analyzed separately vs. the placebo group, then only the
comparison between group B4 and group 0 concerning vertebral BMD
resulted statistical significant (P < 0.05). Adverse events
Only during placebo administration, the following events were reported: headache (two group 0 patients, one group A4 patient, and one group B2 patient), pruritus (one group 0 patient). These disturbances were described as moderate and transient.
During carnitine administration, of the 40 persons in groups A and B, only one A2 patient and one B4 patient complained of disturbances. These consisted in nausea and gastralgia, which had appeared during the first week, disappeared within the next one and were minimal, so that carnitine administration was not discontinued. Transient and moderate gastralgia was also reported by one group 0 patient.
Except for the changes described above for the thyroid hormone-sensitive parameters, in none of the 40 subjects there were significant alterations of erythrocyte sedimentation rate, blood counts (RBC, WBC, platelets), serum electrolytes (Na+, K+, Ca2+, P- ), serum proteins (including protein electrophoresis), bilirubinemia, creatininemia, glycemia, and urine chemistry.
Discussion
Before us only 11 naturally hyperthyroid patients treated with 1
or 2 g/d carnitine for up to 6 weeks had been evaluated (8, 9, 11). Clinical improvement started from the second week and
consisted in the decrease of heart rate, body temperature, asthenia,
nervousness, insomnia, hyperreflexia, and tremors. In a 4-wk study on
iatrogenic hyperthyroidism (10), six male volunteers who
received 100 µg/d L-T3 and 1 g/d
carnitine, were compared with six male volunteers who received only
L-T3. Unfortunately, results are
summarized in a Table where a few parameters are presented (as mean
values without SD or SE) in different ways
(10), thus making interpretation and comparison with our
data impossible. Body weight is reported as mean change, and it was
-10.5 lbs (4.7 kg) or -10.4 lbs (-4.7 kg) in the
T3 group or T3 plus
carnitine group, respectively. Pulse rate is reported as mean final
values (98 or 90 bpm, respectively), and serum cholesterol as mean per
cent change (-38% or -22%) (10). All these studies
(8, 9, 10, 11) failed to specify how symptoms were gauged, and
all were unblinded and devoid of statistical analysis.
With our protocol, we could evaluate whether carnitine is able to both reverse and prevent (or minimize) iatrogenic hyperthyroidism. In groups A, symptoms/signs that had worsened while on L-thyroxine plus placebo returned to baseline once placebo had been substituted with carnitine. In groups B, symptoms/signs remained stable or even ameliorated as long as carnitine was associated to L-thyroxine, implying that the carnitine effect prevailed over the T4 effect. The preponderance of the carnitine effect was confirmed by the negative sign of the second over first bimester change in groups A or the first two bimesters over baseline change in groups B that was observed for a number of biochemical parameters: AST, GGT, SHBG, and ferritin. Therefore, our observations confirm the observations of Hellthaler et al. (7) in rats, in which administration of L-thyroxine + carnitine caused a 10–50% decrease of AST and ALT, while L-thyroxine administration caused a 30–150% increase.
Three biochemical parameters were spared by the antagonism exerted by carnitine on thyroid hormone action: serum TSH, serum osteocalcin, and urinary OH-P. A fourth parameter, circulating total cholesterol is relatively refractory to carnitine, since only the group B showed a statistical difference (P < 0.05) with group 0. The action of carnitine on both TSH (i.e. on thyrotrophs) and OH-P (i.e. on osteoclasts) was nil, because changes matched those caused by placebo. In contrast, the positive response on osteocalcin was greater than that caused by placebo, indicating a synergistic effect with thyroid hormones. In cell cultures, we observed differences among various cell types, because the inhibition of thyroid hormone nuclear uptake was in hepatocytes greater than in neurons (12). Probably, the inhibition of thyroid hormone nuclear entry caused by carnitine in thyrotrophs and osteoclasts is insignificant. Special studies will be required to address this issue, because there are no data available in the literature concerning the carnitine content and uptake in the pituitary and bone. The same ignorance prevents interpretation of the potentiation of the thyroid hormone action on osteoblasts, as indicated by circulating osteocalcin. However, we do confirm the study of Abdennabi et al. (23) in mice. In this study (23), three groups of five adult mice each were tested with placebo or 50 mg/kg or 100 mg/kg body weight oral carnitine for 12 wk. In the last two groups, serum osteocalcin increased by 22% and 60% compared with the placebo group.
Because of the different effect of carnitine on osteoclasts and osteoblasts, we should expect a beneficial effect on bone, although of low magnitude considering the relatively short period of administration of carnitine and the consequent interference with only one bone formation cycle. The beneficial effect of carnitine can be appreciated better in the B groups, because they received carnitine for a longer period compared with the A groups. Even though in groups B the posttreatment BMD was measured at the sixth month, namely 2 months after withdrawal of carnitine, there was an average 1.8% (group B2) and 2.3% (group B4) increment in lumbar spine BMD, and a 1.0% (B2) and 1.3% (B4) increment in femur BMD. These beneficial changes contrasted with the -0.6% and +0.2% changes observed in the placebo group. Because of the short duration of treatment and small number of patients as a result of a dropout, statistically significant (P < 0.05) was only the difference in vertebral BMD between group 0 and group B4. Clearly, if the increments of BMD are confirmed on a larger series of patients treated for a longer period of time, the adjunctive carnitine therapy is ideal for postmenopausal women who need to take life-long TSH-suppressive doses of L-thyroxine.
When iatrogenic hyperthyroidism appears, several patients spontaneously reduce the dose of L-thyroxine or take L-thyroxine irregularly or may even stop therapy. In each case, the final result is the absence of consistent suppression of TSH—and this is deleterious in thyroid carcinoma patients. The traditional approach to avoid hyperthyroidism is to individually tailor the dose of L-thyroxine, but this requires frequent clinical and hormone evaluations. (Even by doing so, some patients continue to complain of side effects, because the individual "euthyroid" range is narrower than the population range.) These frequent controls can be eliminated by using a fixed TSH-suppressive daily dose of L-thyroxine plus L-carnitine at 2 g per day or, as preliminary data of ours indicate, even 1 g per day. Alternatively, carnitine can be added only after hyperthyroid symptoms had appeared. After our clinical trial ended and results were known, patients were informed of the cost of carnitine and inquired. Group A and B patients were inquired if they were willing to continue the adjunctive therapy with 2 g carnitine per day; group 0 patients were asked if they were willing "to take twice daily and orally 2 g of a medication, as natural as thyroxine, that would have protected them from the side-effects of the hormone." The proportion of the favorable responders (34 of 40 or 85% and 8 of 10 or 80%) were similar. When the 42 favorable responders were asked if they preferred to take carnitine for only a strictly limited period of time or for as long as necessary, 33 (79%) preferred the second modality.
On a daily basis, the cost of 1 g carnitine is $0.65, which
compares nicely with benzodiazepines (e.g. $0.42 to $0.6 for
3 mg bromazepam or 2 mg lorazepam) or ß-blockers (e.g.
$0.1 for 60 mg propanolol), considering that these drugs act on select
symptoms and have significant side effects. Other drugs that have been
used to counteract hyperthyroidism are bile sequestrants
(24, 25, 26), the cons of which are the high daily cost
($5 for 8 g cholestyramine or 20 g colestipol), the high
frequency of gastrointestinal disturbances, the impaired absorption of
fat-soluble vitamins and several drugs, and the main effect being
limited to the first week of treatment (26).
Theoretically, other drugs that interfere with thyroid hormone either
transport accross cells or interaction with nuclear receptors
(e.g. diphenylhydantoin, nonsteroidal antiinflammatory
drugs, amiodarone, etc.) (27) might also be
used. However, none of all the above drugs is naturally occurring, and
each has important side-effects. In sharp contrast, carnitine has no
toxicity, teratogenicity, known contraindications, interaction with
drugs, or important side effects. In addition, hyperthyroidism
impoverishes the tissue deposits of carnitine (28, 29),
thus creating a true situation of secondary carnitine deficiency
(4). In our group 0 patients, 6 months of
L-thyroxine therapy increased the urinary
excretion of carnitine by almost 3-fold (Fig. 1), reflecting its
thyroid hormone-induced efflux from cells. If we also consider that
carnitine inhibits T3 and
T4 entry into the nucleus of a number of
peripheral cells (12), then there is a double rationale
for the use of carnitine at least as ancillary therapy of
hyperthyroidism: to replenish the tissue deposits and to counteract
thyroid hormones in the periphery.
Carnitine could be of particular use when it is important to use the lowest possible dose of antithyroid drugs, such as pregnancy, lactation and liver and/or hematologic disorders (30, 31), since these drugs cross the placenta, are secreted into the milk, and have liver and hematopoietic toxicity (14, 30, 31). Because carnitine crosses the placenta and is devoid of fetal toxicity (32), and because of the difficulties to treat fetal thyrotoxicosis (33)—a serious complication of a number of pregnant women with Graves’ disease—carnitine would be an ideal drug in this setting. Also forms of thyrotoxicosis due to leakage of thyroid hormones, or antithyroid drug-resistant thyrotoxicosis (e.g. amiodarone-related thyrotoxicosis) or thyroid storm are amenable to treatment with carnitine. Thyroid storm is a serious emergency that has a mortality of 20–50% and which is triggered by precipitating events (34). Thus, carnitine would be useful both for the prevention and the therapy of the thyroid storm.
Acknowledgments
We are indebted to Dr. Nadia C. Aricò for the statistical analysis of the data and to Sigma-Tau (Pomezia, Italy), particularly Dr. Carlo Trevisani, for the supplies of carnitine. We also thank Drs. A. Toscano and M. D’Aghennouz (Clinica Neurologica 2, Messina, Italy) for the assay of urinary carnitine. Drs. L. Bartolone and S. Squadrito participated in the early stages of this work.
Footnotes
Abbreviations: ALT, alanine aminotransferase; AST,
aspartate aminotransferase; BMD, bone mineral density; CI, confidence
interval; CV, coefficient(s) of variation; FT3, free
T3; FT4, free T4; GGT,
-glutamiltransferase; OH-P hydroxyproline; SHBG, sex hormone-binding
globulin.
Received September 25, 2000.
Accepted April 18, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  T. Watt, M. Groenvold, A. K. Rasmussen, S. J. Bonnema, L. Hegedus, J. B. Bjorner, and U. Feldt-Rasmussen Quality of life in patients with benign thyroid disorders. A review. Eur. J. Endocrinol., April 1, 2006; 154(4): 501 - 510. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. MANOLI, M. U. DE MARTINO, T. KINO, and S. ALESCI Modulatory Effects of l-Carnitine on Glucocorticoid Receptor Activity Ann. N.Y. Acad. Sci., November 1, 2004; 1033(1): 147 - 157. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. BENVENGA, A. AMATO, M. CALVANI, and F. TRIMARCHI Effects of Carnitine on Thyroid Hormone Action Ann. N.Y. Acad. Sci., November 1, 2004; 1033(1): 158 - 167. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. ALESCI, M. U. DE MARTINO, T. KINO, and I. ILIAS l-Carnitine Is a Modulator of the Glucocorticoid Receptor Alpha Ann. N.Y. Acad. Sci., June 1, 2004; 1024(1): 147 - 152. [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 |
). In group B2 (n = 10)
and B4 (n = 10) patients, who were enrolled to evaluate whether
2–4 g/d oral L-carnitine were able to prevent iatrogenic
hyperthyroidism, placebo was associated during months 5 and 6, and
carnitine during months 1 through 4 (
