Me. Toubert et al., Cervicomediastinal magnetic resonance imaging in persistent or recurrent papillary thyroid carcinoma: Clinical use and limits, THYROID, 9(6), 1999, pp. 591-597
Cervicomediastinal magnetic resonance imaging (MRI) was evaluated in 13 con
secutive persistent or recurrent papillary thyroid carcinoma (PTC) patients
, previously treated by total thyroidectomy and radioiodine ablation. All h
ad elevated thyroglobulin (Tg) levels and were therefore submitted to a new
therapeutic radioiodine dose followed by a posttherapeutic whole-body scan
(I-131-WBS),,and subsequent MRI. Patients with known distant metastases we
re excluded from the study. Group 1 included 7 patients with a negative I-1
31-WBS, whereas cervical and/or mediastinal I-131-uptake was evidenced in t
he other 6 patients (group 2). MRI was thus compared to I-131-WBS, and addi
tionally in 8 reoperated cases, to histology. MRI was positive in 11 of 13
(85%) patients, corresponding to 23 of 55 (41.8%) histologically confirmed
sites. In group 1, MRI was positive in 5 of 7 patients, with a sensitivity
of 47% (15/32 histologically positive sites), allowing appropriate indicati
on of surgery: 4 neck surgery, and 1 mediastinal dissection because of too
distant lymph node foci. In group 2, MRI always showed more localization th
an I-131-WBS; histology was obtained in 3. Because all the foci located in
the mediastinal area (0.8 to 1.8 cm) were histologically confirmed (7/7 sit
es), MRI avoided underestimation of surgery in the 8 reoperated patients. H
owever, additional images were also observed corresponding to a normal thym
us, a small neuroma or inflammatory lymph nodes, but pretracheal and very s
mall nodes (less than 0.5 cm) were missed. In conclusion, although less spe
cific than radioiodine scintigraphy, MRI can detect local persistent or rec
urrent PTC, and seems particularly effective for evaluation of mediastinal
involvement.