versão impressa ISSN 0100-3984
Radiol Bras v.39 n.4 São Paulo jul./ago. 2006
Sclerosis: an option for treatment of cystic thyroid nodules*
Daysi Maria de Alcântara-JonesI; Leila Maria Batista AraújoII; Alessandro de Moura AlmeidaIII; Daniel de Alcântara JonesIV; Julia Mandaro LavinasIV; Lázaro José Góes CardosoV; Marize Carvalho PassosV
Course of Post-Graduation in Medicine & Health, Universidade Federal da
IIAdjunct Doctor Professor at Department of Internal Medicine, Universidade Federal da Bahia
IIIAcademic of Medicine, Universidade Federal da Bahia
IVAcademics of Medicine, Universidade Severino Sombra, Rio de Janeiro, RJ, Brazil
VPhysicians at Service of Radiodiagnosis, Hospital São Rafael
To check the effect of ultrasound-guided percutaneous ethanol injection for
treatment of cystic thyroid nodules.
MATERIALS AND METHODS: A comparison was made of the volume of 34 benign cystic nodules in 30 patients (26 female, four male), before and 3.9 months (on average) after alcoholization. For the purpose of finding out interobserver variations in echographic measurements, the initial volume of the nodules was assessed by means of ultrasound at different moments, by two observers, without any previous knowledge on the nodules size.
RESULTS: Mean volume of nodules before treatment was 12.3 ± 18.0 ml. Mean reduction rate in nodules after alcoholization was 74.0 ± 26.1% (p < 0.0001) and 7/34 (20.6%) of the nodules disappeared. There was no correlation between nodules size and reduction rate. The average interobserver nodule measurement variation was 0.5 ml, for an alpha of 5%. Moderate pain at the time of injection was the most frequent complication.
CONCLUSION: Ultrasound-guided percutaneous ethanol injection is a safe and effective option for treatment of cystic thyroid nodules.
Keywords: Thyroid gland nodule; Nodular goiter; Ethanol; Interventional ultrasonography; Prospective studies.
It is known that about 5% to 7% of population have a thyroid gland nodule and less than 10% of them are malignant, determining surgical indication(14). Both conventional and conservative forms of benign thyroid nodules treatment have been frustrating in terms of reduction and avoidance of the resulting local compression or even esthetical phenomena(5).
Percutaneous ethanol injection (PEI) was initially utilized for treating primary hyperparathyroidism, resulting in destruction of hyperplastic parathyroids(6) and for reducing tumor masses(7,8). Later, the same group of radiologists started to use this procedure with success for treating thyroid nodules. Early in the nineties, Livraghi et al.(9) achieved a significant reduction in volume of eight "hot" and autonomous thyroid nodules, besides obtaining the resolution of hyperthyroidism in these patients. Additionally, other authors have confirmed the effectiveness of this method in the treatment of hyperfunctioning thyroid nodules(1015) and in "cold" nodules(1618), but cystic lesions are best-responsive to the alcohol sclerosing effects(1922).
In the United States, less than 8% of clinicians indicate this procedure(23). In Brazil, there are few studies published on the subject(24,25) reporting experiences of thyroid nodules treatment by means of alcoholization. Therefore, the objective of this study was to evaluate the PEI effectiveness in reducing cystic thyroid nodules, comparing the evolution of such nodules before and after alcoholization.
MATERIALS AND METHODS
Thirty patients presenting uninodular and multinodular goiters were selected in the ambulatory of Hospital São Rafael, Salvador, BA, Brazil, in the period between May/2001 and April/2005, with a total of 34 nodules studied. For this treatment cystic or predominantly cystic lesions were elected. The patients had either a solitary thyroid nodule, or a nodule in each thyroid lobe, or an unquestionably largest dominating nodule, aiming at evaluating each nodule evolution without confounding it with other nodule that was not being treated. There was no restriction to the thyroid nodules size.
This study was approved by the Hospital São Rafael Research Ethics Committee. All the patients have signed an informed and free consent, during an interview, after receiving all the necessary information on the technique.
All the patients underwent clinical and laboratory tests for determining levels of T3, free T4, ultra-sensitive TSH and antithyroid peroxidase antibody. The patients were not using any medication interfering with the thyroid function since the very moment they started being observed with the first ultrasound (US). No patient had a family history of thyroid cancer, neither had he/she been previously submitted to thyroid surgery.
All the thyroid nodules underwent US-guided fine needle aspiration biopsy in their solid portion or on the cystic wall to rule out malignancy. The nodules were emptied and the whole aspirate fluid was sent for cytological analysis. Nodules that presented at standard cytology with cells overlapping or predominance of typical or atypical Hürthle cells, or even presenting echographic aspects suggesting malignancy, were excluded from the study, as previously described(26).
US studies were performed with an Aloka SSD 1700 color Doppler model, using a 7.5 MHz transducer. The distances (longitudinal, transverse and anteroposterior) between the points more distant from the center of the nodule were determined, the product of these three dimensions being multiplied by a constant (3.14/6) to estimate the thyroid nodule volume before and after alcoholization(24).
The initial volume of the nodules was evaluated with US by two observers at different moments, without any previous information on their size, aiming at knowing the interobserver variation between echographic measurements. Nodules with > 10 ml could not be visualized on the US device monitor in just one image capture and, since the differences between measurements provided by the observers were quite significant, only thyroid nodules measurements (n = 25) minor or equal to 10 ml were considered for calculation of these variations.
At each new phase of the treatment, the patient had the volume of his/her thyroid nodule measured after the later alcoholization, the observer not being informed on the previous measurement results.
The ultrasound standard has allowed the recognition of two types of cystic lesions: the predominantly cystic lesions (fluid component greater than 50%) and simple cystic lesions (solid portion minor than 10%).
The PEI therapy was performed with the patient lying in dorsal decubitus position, with a pillow placed under the scapular region, aiming at producing neck hyperextension. The procedure has not demanded anesthesia. Immediately before the PEI, the nodule fluid content was aspirated, as recommended(19,27), with a 10 ml syringe and a 30 mm x 8 mm needle, taking care to leave part of the fluid in the cavity in order not to loose the needle visualization. Once the aspirate had been obtained, the syringe was disconnected from the needle and the aspirate was discharged. With the needle still inserted, a new syringe loaded with the alcohol was adapted. In every cases the intention was injecting an ethanol amount equivalent to 50% of the fluid aspirate volume(20), although at that moment the patient's acceptance was the determining factor of the effectively injected amount of alcohol(19,20). After a 15-minute permanence of the alcohol in the cavity, part of the volume injected was aspirated back into the syringe.
The treatment was considered completed when the thyroid nodule disappeared or became very small, complicating its puncture by the needle. The absence of vascularization inside and on the cyst walls and ethanol injection in an amount equivalent of superior to the thyroid nodule volume also were situations in which the treatment was considered as completed. On the other hand, the treatment was considered as interrupted in cases where it was not well accepted by the patient.
Each patient had only one thyroid nodule treated per session. Also, according their reduction degree, the thyroid nodules were classified into: non-responsive (< 20% reduction), partially responsive (> 20% and < 50% reduction) and responsive (> 50% reduction).
The continuous variables were described by mean ± standard deviation (SD) and the categorical variables, by absolute numbers and ratios. The thyroid nodule reduction rate was given by the ratio between volume differences (final initial) and initial volume, multiplied by 100.
The Wilcoxon test was applied to compare thyroid nodules initial and final volumes. The Mann-Whitney test compared reduction rates of two groups of ultrasound standards (cystic and predominantly cystic nodules). The chi-square test compared the groups (responsive, partially responsive and non-responsive).
The Spearman correlation test evaluated the correlation of volume reduction rate with initial nodule volume, aspirate fluid volume and volume of ethanol injected into the thyroid nodule.
The Bootstrap method(28) verified the significance of the difference between the thyroid nodules volume measurements at US provided by the observers for a 5% alpha.
The calculations were performed with the Statistical Package for Social Sciences, version 11.5 for Windows and Ox 3.30 (Oxford), p values less than 5% (p < 0,05) being considered as statistically significant.
Thirty patients (four male, 26 female) presenting uni- or multinodular goiters were submitted to alcoholization, in a total of 34 cystic thyroid nodules treated. Demographic, clinical and laboratory characteristics are shown in Table 1.
As regards their ultrasound standard, the thyroid nodules were rated into: cystic in 38.2% (n = 13) and predominantly cystic in 65.0% (n = 21).
A significant variation between thyroid nodules initial volume (0.3 to 82.7 ml, mean ± SD 12.3 ± 18.0 ml, median 4.8 ml) and final volume (0.0 to 10.3 ml, mean of 2.2 ± 2.7 ml, median of 1.1 ml) was observed, and this variation was statistically significant (p < 0.0001) (Graphic 1). The mean reduction was 74.0 ± 26.1%. The mean follow-up time after treatment was 3.9 ± 4.7 months (1.023.0).
The cystic thyroid nodule content aspiration was satisfactory in all of the cases, with the exception of one, where a thick fluid obstructed the needle. A lesser amount of ethanol was injected and the resulting reduction (45.9%) was much inferior to the registered mean. In the other cases, after ethanol injection, a three to five-minute interval was allowed to aspirate the cyst for the second time. Usually, at the second session, the thyroid nodule was sensitively smaller (a more than 50% reduction) and the ultrasound standard was hyper- or isoechogenic.
The mean aspirate fluid volume in the first session was 18.7 ± 20.3 ml and mean injected ethanol amount during the whole treatment was 7.2 ± 10.0 ml. Fourteen thyroid nodules underwent alcoholization in only one session and 18 in two sessions. The latest has a higher reduction rate than those submitted to only one session (p < 0.0001). Two thyroid nodules, a cystic one and another predominantly cystic whose fluid content could not be aspirated, did not present any reduction after the first alcoholization session. In the second and third sessions, the fluid content was punctured with a 30 x 12 mm gauge needle, resulting in a large amount of a thick fluid. In one of them, whose fluid was thicker the reduction rate was 52.9%, and this was the only nodule to be alcoholized in four sessions.
Only three (8.8%) nodules were considered as non-treatment-responsive, four (11.8%) partially responsive and 27 (79.4%), responsive, and of these nodules, seven (20.6% of the total) disappeared. The difference between groups was statistically significant (p < 0.00001). Figure 1 shows an echographic image of an 83-year-old man who presented dry cough for several months. A 55.4 ml thyroid cyst was diagnosed. After an alcoholization session, the cough has resolved and the nodule has substantially decreased. An US scan performed seven months later (Figure 2) has revealed thyroid nodule absence.
Additionally, one has observed that the greater the nodule cystic content, the larger the volume reduction rate (p < 0.006), as per Graphic 2. There was no statistically significant for correlation between reduction rate and initial nodule volume (r = 1.0; p = 0.6). A nodule reduction rate was not correlated with its initial volume (r = 1.0; p = 0.6), nor with its fluid aspirate volume (r = 0.6; p = 0.08), nor with the injected ethanol volume (r = 0.2; p = 0.3).
In 25 thyroid nodules smaller or equal to 10 ml, two measurements were performed by two independent sonographists and the significance of the difference between these measurements was calculated for a 5% alpha. The mean difference between measurements was 0.5 ml and the likelihood of a > 0.7 difference being due to change is less than 5%.
In the 57 alcoholization sessions performed, the most frequent complication was pain (55.9%), most of times ob mild (23.6%) and moderate (23.6%) intensities (Table 2). Frequently, the pain irradiated generally to the homolateral ear and to the dental arch. In 49.1% (n=28) of cases, patients have not reported any kind of discomfort. In the cases where pain was reported, this only manifested at the moment of injection.
The treatment was considered as completed in 27 (79.4%) cases, because in seven (20.6%) the nodules disappeared; in ten (29.4%) the nodules became too small to be punctured, and in three (8.8%) started to present an increase in pressure inside the nodule, making injection difficult. After treatment, 7/34 (20.6%) of thyroid nodules became hyper- or isoechogenic and without any vascularization. The treatment was interrupted in 20.6% of cases (two patients presented a lower-than-expected reduction in the first session and the authors lost contact with five patients).
In the present study, one has observed a thyroid nodules mean reduction of 74 ± 26%, superior to the reduction found by Wesche et al.(29), who utilized radioactive iodine in a group of 32 patients presenting nodular goiters (46% reduction) compared to a group (n = 32) using dose-suppressive thyroid hormone therapy for two years (22% reduction rate). It is important to note that, in the latest group, 57.1% of nodules have not presented any response to the treatment.
Our results are similar to those of Bannedbaek and Hegedüs, who applied PEI in 33 bulky cystic lesions (median 8 ml) presenting compression symptoms; after three months, 82% of them measured less than 1 ml, most of them undergoing only one alcoholization session(20).
According other authors criteria(20,24), the data of this series demonstrate that the thyroid nodules responded to treatment with a reduction rate higher than 20% of their initial volume, with one or two alcoholization sessions, in 94.1% of cases. These data reflect higher rates than those of Lima et al.(5), who attempted to reduce solid thyroid nodules with a mean of 1.7 ± 2.5 ml during 12 months, with TSH suppression therapy. Only 11/54 (20.3%) of patients achieved a reduction rate higher than 20%. These data demonstrate the superiority of PEI in relation to the TSH suppression therapy.
In this series, the thyroid nodules volumetric reduction percentage also was higher than that found in other series utilizing the treatment in nodules of different ultrasound standards or even those utilizing the study in solid thyroid nodules. Alcântara-Jones et al.(30), performing one or two alcoholization sessions in 84 thyroid nodules of different ultrasound standards, found a cystic thyroid nodules reduction rate of 90.95% while solid nodules reduced 37.3%.
The present study includes information not published by other authors up to this moment and that may constitute a limitation to the use of PEI for treating cystic thyroid nodules: cysts with a very thick content may not present any reduction with sclerosis by means of alcoholization. It is important to note that the predominantly cystic nodule whose fluid could not be aspirate in the first session, presented a 26.0% increase in volume. In the second session, a reduction was already observed, although inferior to that presented by other cystic lesions, considering a same number o treatment sessions. Therefore, it may be concluded that, with a diagnostic puncture to rule out malignancy, we can reach a more accurate evaluation of the possibility or not of success with alcoholization.
Several studies evaluating thyroid nodules reduction either utilizing PEI(18) or conventional treatment with suppressive doses of thyroid hormone(5,29), demonstrate that smaller thyroid nodules decrease more easily than the larger ones. In this series, no correlation was found between the reduction rate and the initial nodule volume.
There was a perfect correlation between reduction rates and initial thyroid nodules volume, but, possibly because of the small number of nodules studied, we could not achieve statistical significance. Also, no significant correlation was found between the thyroid nodule reduction rate and the volume of aspirate fluid from cyst-like lesions. This was an expectation of the observers who frequently noted the easiest reduction of cystic and cyst-like nodules from which a larger amount of fluid was aspirate before alcoholization. The explanation for the failure in determining a significant correlation is the fact that we had in our sample only ten cystic nodules and possibly because we were insufficiently aggressive in aspirating the fluid before alcoholization.
Along the evaluation of non-surgical methods of thyroid nodules treatment, it is necessary to know the errors in measurement of thyroid nodules, especially when the structures studied are larger than 12 ml. This concern has been highlighted by some authors(31,32).
The interobserver variation coefficient for repeated measurements of a thyroid nodule in a same patient, obtained in 25 randomly selected patients, was 0.7 ml. This difference was similar to that obtained by Celani(33), who has found a mean discordance of 0.5 ± 0.03 ml in 40 determinations and lower than the 11.7 difference found by Papini et al., who have accounted measurements performed by three observers in a 100-patient population(31), with thyroid nodules presenting a mean volume of 1.53 ± 0.6 ml. Another information provided by this evaluation is that, if a thyroid nodule presents a 0.7 ml decrease in volume between sessions, this is due to an actual change in volume and not to a variation in measurements related to the examination.
The low interobserver variation coefficient, also registered in this series, suggests that US is a reliable method for following-up patients undergoing this treatment. However, it may be observed that the average aspirate fluid amount in the first session was higher than the average initial nodules volumes, which allows the conclusion that the measurements provided by the observers have underestimated the actual thyroid nodules volume and possibly this has occurred among larger than 10 ml nodules that have not been taken into consideration for this calculation.
This study was limited by the variation in thyroid nodules sizes, a difficulty that has been presented by other authors(16,24,25) who also deal with thyroid nodules heterogeneity in relation to several aspects, including the response to this kind of treatment.
The reference to pain as the main complication of this technique has been reported in several series and usually it is described as transitory, generally lasting for less than ten minutes, corresponding to the duration of a sclerosis session(20). There are reports from other author who have utilized xylocaine associated with ethanol in sclerosis(34), with an improvement in relation to this complaint, although this has not produced better outcomes in other series(24).
This study demonstrates that, differently from the natural evolution of thyroid nodules, PEI results in an effective reduction of thyroid nodules dimensions. It is necessary to follow-up the medium- and long-term evolution of these lesions, in order to allow an actual dimensioning the effectiveness of this therapeutic modality.
We have concluded that PEI is an alternative, effective and safe method for treatment of thyroid nodules, the higher lesion cystification degree resulting in a higher efficiency of the method, whose complications are transitory and autolimited.
1. Mazzaferri EL. Management of a solitary thyroid nodule. N Engl J Med 1993;328:553559. [ Links ]
2. Samuels MH. Evaluation and treatment of sporadic nontoxic goiter some answers and more questions. J Clin Endocrinol Metab 2001;86: 994997. [ Links ]
3. Murussi M, Pereira CEFN, Brasil BMAA, Tschiedel B. Punção aspirativa da tireóide com agulha fina em um hospital geral: estudo de 754 punções. Arq Bras Endocrinol Metab 2001;45:576583. [ Links ]
4. Lima MA, Yamada AF, Navarro FC, Resende EM, Ferreira B, Borges MF. Punção biópsia aspirativa de tireóide em região endêmica de bócio colóide. Arq Bras Endocrinol Metab 2002;46:275279. [ Links ]
5. Lima N, Knobel M, Cavaliere H, Sztejnsznajd C, Tomimori E, Medeiros-Neto G. Levothyroxine suppressive therapy is partially effective in treating patients with benign, solid thyroid nodules and multinodular goiters. Thyroid 1997;7:691697. [ Links ]
6. Solbiati L, Giangrande A, De Pra L, Bellotti E, Cantù P, Ravetto C. Percutaneous ethanol injection of parathyroid tumors under US guidance: treatment for secondary hyperparathyroidism. Radiology 1985;155:607610. [ Links ]
7. Livraghi T, Festi D, Monti F, Salmi A, Vettori C. US-guided percutaneous alcohol injection of small hepatic and abdominal tumors. Radiology 1986;161:309312. [ Links ]
8. Livraghi T, Salmi A, Bolondi L, et al. Small hepatocellular carcinoma: percutaneous alcohol injection results in 23 patients. Radiology 1988;168: 313317. [ Links ]
9. Livraghi T, Paracchi A, Ferrari C, et al. Treatment of autonomous thyroid nodules with percutaneous ethanol injection: preliminary results. Work in progress. Radiology 1990;175:827829. [ Links ]
10. Goletti O, Monzani F, Caraccio N, et al. Percutaneous ethanol injection treatment of autonomously functioning single thyroid nodules: optimisation of treatment and short-term outcome. World J Surg 1992;16:784790. [ Links ]
11. Livraghi T, Paracchi A, Ferrari C, Reschini E, Macchi RM, Bonifacino A. Treatment of autonomous thyroid nodules with percutaneous ethanol injection: 4-year experience. Radiology 1994; 190:529533. [ Links ]
12. Lippi F, Ferrari C, Manetti L, et al. Treatment of solitary autonomous thyroid nodules by percutaneous ethanol injection: results of an Italian multicenter study. The Multicenter Study Group. J Clin Endocrinol Metab 1996;81:32613264. [ Links ]
13. Cerbone G, Spiezia S, Colao A, et al. Percutaneous ethanol injection under power Doppler ultrasound assistance in the treatment of autonomously functioning thyroid nodules. J Endocrinol Invest 1999;22:752759. [ Links ]
14. Tarantino L, Giorgio A, Mariniello N, et al. Percutaneous ethanol injection of large autonomous hyperfunctioning thyroid nodules. Radiology 2000;214:143148. [ Links ]
15. Zingrillo M, Torlontano M, Ghiggi MR, et al. Radioiodine and percutaneous ethanol injection in the treatment of large toxic thyroid nodule: a long-term study. Thyroid 2000;10:985989. [ Links ]
16. Bennedbaek FN, Nielsen LK, Hegedüs L. Effect of percutaneous ethanol injection therapy versus suppressive doses of L-thyroxine on benign solitary solid cold thyroid nodules: a randomized trial. J Clin Endocrinol Metab 1998;83:830835. [ Links ]
17. Goletti O, Monzani F, Lenziardi M, et al. Cold thyroid nodules: a new application of percutaneous ethanol injection treatment. J Clin Ultrasound 1994;22:175178. [ Links ]
18. Zingrillo M, Collura D, Ghiggi MR, Nirchio V, Trischitta V. Treatment of large cold benign thyroid nodules not eligible for surgery with percutaneous ethanol injection. J Clin Endocrinol Metab 1998;83:39053907. [ Links ]
19. Verde G, Papini E, Pacella CM, et al. Ultrasound guided percutaneous ethanol injection in the treatment of cystic thyroid nodules. Clin Endocrinol (Oxf) 1994;41:719724. [ Links ]
20. Bennedbaek FN, Hegedüs L. Treatment of recurrent thyroid cysts with ethanol: a randomized double-blind controlled trial. J Clin Endocrinol Metab 2003;88:57735777. [ Links ]
21. Guglielmi R, Pacella CM, Bianchini A, et al. Percutaneous ethanol injection treatment in benign thyroid lesions: role and efficacy. Thyroid 2004; 14:125131. [ Links ]
22. Lee SJ, Ahn IM. Effectiveness of percutaneous ethanol injection therapy in benign nodular and cystic thyroid diseases: long-term follow-up experience. Endocr J 2005;52:455462. [ Links ]
23. Bonnema SJ, Bennedbaek FN, Ladenson PW, Hegedüs L. Management of the nontoxic multinodular goiter: a North American survey. J Clin Endocrinol Metab 2002;87:112117. [ Links ]
24. Braga-Basaria M, Trippia MA, Stolf AR, Mesa Jr C, Graf H. Tratamento de nódulos autônomos e císticos da tireóide com injeção intranodular de etanol. Rev Assoc Med Bras 2002;48:335340. [ Links ]
25. Bianchini EX, Ikejiri ES, Mamone MC, Paiva ER, Maciel RMB, Furlanetto RP. Injeção percutânea de etanol no tratamento de nódulos tiroidianos sólidos, císticos e autônomos. Arq Bras Endocrinol Metab 2003;47:543551. [ Links ]
26. Camargo RYA, Tomimori EK. Diagnóstico dos nódulos tireóideos baseado na avaliação ultra-sonográfica e citológica combinada. Arq Bras Endocrinol Metab 1998;42:273276. [ Links ]
27. Iacconi P, Spinelli C, Monzani F, Miccoli P. Percutaneous ethanol injection for thyroid cysts: a word for caution. Clin Endocrinol (Oxf) 1996;44: 126. [ Links ]
28. Efron B, Tibshirani RJ. An introduction to the Bootstrap. Chapman & Hall/CRC, 1993. [ Links ]
29. Wesche MFT, Tiel-V Buul MMC, Lips P, Smits NJ, Wiersinga WM. A randomized trial comparing levothyroxine with radioactive iodine in the treatment of sporadic nontoxic goiter. J Clin Endocrinol Metab 2001;86:9981005. [ Links ]
30. Alcântara-Jones DM, Araújo LMB, Almeida AM, Jones DA, Cardoso LJG, Passos MC. Efeito da injeção percutânea de etanol na redução de nódulos tireoideanos. Arq Bras Endocrinol Metab 2006;50:97104. [ Links ]
31. Papini E, Petrucci L, Guglielmi R, et al. Long-term changes in nodular goiter: a 5-year prospective randomized trial of levothyroxine suppressive therapy for benign cold thyroid nodules. J Clin Endocrinol Metab 1998;83:780783. [ Links ]
32. Zelmanovitz F, Genro S, Gross JL. Suppressive therapy with levothyroxine for solitary thyroid nodules: a double-blind controlled clinical study and cumulative meta-analyses. J Clin Endocrinol Metab 1998;83:38813885. [ Links ]
33. Celani MF. Levothyroxine suppressive therapy in the medical management of nontoxic benign multinodular goiter. Exp Clin Endocrinol 1993;101: 326332. [ Links ]
34. Papini E, Panunzi C, Pacella CM, et al. Percutaneous ultrasound-guided ethanol injection: a new treatment of toxic autonomously functioning thyroid nodules? J Clin Endocrinol Metab 1993;76: 411416. [ Links ]
Dra. Daysi Maria de Alcântara-Jones
Loteamento Jardim do Atlântico, Quadra C, Lote 19
Lauro de Freitas, BA, Brazil 42700-000
Accepted after revision December 12, 2005.
* Study developed at Services of Endocrinology and Radiodiagnosis, Hospital São Rafael Monte Tabor Centro Ítalo-Brasileiro de Promoção Sanitária, Salvador, BA, Brazil.