Comparison of Outcomes After ¹²³I Versus ¹³¹I Preablation Imaging Before Radioiodine Ablation in Differentiated Thyroid Carcinoma

MATERIALS AND METHODS

Fifty consecutive patients with histopathologically proven papillary or follicular thyroid carcinoma without detectable distant metastases were studied after total thyroidectomy when a minimal blood level of serum thyroid-stimulating hormone (TSH, thyrotropin) of 30 μIU/mL had been achieved. The patients were randomly assigned to 1 of 2 groups. Group 1 (n = 26) received 14.8 ± 0.2 MBq (mean ± SD) of ¹²³I (as sodium iodide) 24 h before imaging, with ablative therapy given the next day, using 3.7 GBq of ¹³¹I (as sodium iodide) if the ¹²³I scan showed only a postthyroidectomy remnant. Group 2 patients (n = 23) were dosed with 74.0 ± 0.3 MBq of ¹³¹I and scanned at 24, 28, and 72 h; they were then treated at 72 h with the same empiric ablative dosage of 3.7 GBq of ¹³¹I if the scan showed a postthyroidectomy remnant without distant metastatic disease. The 72-h delay was required because of a quantitative assessment performed for another study that did not affect the patient's treatment in any way. Follow-up scintigraphy was performed in the same hypothyroid state 6−8 mo after therapy, using 74 ± 0.3 MBq of ¹³¹I. Successful ablation of the initial thyroid remnant was defined as both the absence of visible iodine-concentrating tissue in the thyroid bed and adjacent lymph nodes on this follow-up scan and a simultaneous serum tg level below the lower limit of detection in the presence of a negative assay for anti-tg antibodies. Statistical comparisons of groups were made by χ² analysis with a P value required to be <0.05 to indicate statistically significant differences. The study was performed under the guidelines of the University of Cincinnati Institutional Review Board.

RESULTS

Fifty patients were entered into the study, with one (from group 2) lost to follow-up. The patients of the 2 groups were found to be statistically comparable in age, sex ratio, ratio of the number of papillary to follicular thyroid carcinomas, tumor burden, and locoregional metastases. None had distant metastases (Table 1). The mean activity of the ablative dosage of ¹³¹I was not significantly different between the 2 groups, averaging 3.77 ± 0.11 GBq.

These demographically similar groups had comparable scintigraphic rates of ablation (group 1, 88%; group 2, 91%; P > 0.05), but, using our second diagnostic criterion for complete ablation—that the quantitative measurement of serum tg had to be below detectable limits as well (unlike all but one of the studies to be discussed)—decreased our ablative success rate for group 1 to 81% and for group 2 to 74% (P > 0.05) (Table 2). Six patients in group 1 (23%) and 4 in group 2 (17%) (P > 0.05) had an elevated anti-tg antibody level, invalidating the serum tg test as a criterion of the completeness of ablation for these 10 patients.

DISCUSSION

We have demonstrated that if stunning from ¹³¹I occurred in our patients, as much of the relevant literature suggests, this phenomenon did not reduce the yield of complete ablation of thyroid remnants when compared with a statistically similar sample of patients we studied who received ¹²³I as the diagnostic radionuclide. For the detection of residual normal tissue (not thyroid cancer), the ¹²³I scan performed at 24 h is quite sensitive at the dosage we used.

We believe that a preablation radioiodine scan is important for several reasons: (a) about 2% of our patients have had a true total thyroidectomy by both scintigraphic evidence and the absence of measurable serum tg, and it is not justifiable to administer β-emitting radioiodine therapy to patients who do not need it; (b) staging with the preablation iodine dosage may reveal metastatic disease, a finding that then requires a higher therapeutic amount of iodine than originally planned; (c) some areas of iodine-concentrating metastatic disease may require special preparations—for example, corticosteroids before treating cerebral metastases.

Several problems became apparent on review of the relevant literature:

• (i) The comparison groups were not always matched for age, sex, type of thyroid cancer, and diagnostic and therapeutic dosages of radioiodine;

• (ii) If stunning was noted, it was usually confirmed only qualitatively, a process which can cause its own problems (4);

• (iii) The great majority of papers in this field have judged the success of ablation solely from visual inspection of a follow-up scan 6 or more months after therapy with ¹³¹I; however, one must also require a negative assay for serum tg to exclude the presence of functioning thyroid tissue.

Stunning, as defined, was first demonstrated in 1951 (5), and this phenomenon has subsequently been reported in numerous papers and summarized in recent reviews (6,7). An inference from the observation of stunning is that the diagnostic dosage has interfered in some way with the trapping or retention of the therapeutic dosage or that the targeted mass has been reduced in volume and, hence, in uptake.

Numerous studies indicate that ¹²³I is comparable to ¹³¹I in the detection of thyroid remnants after thyroidectomy (8–10). Park et al. studied patients after thyroidectomy with diagnostic dosages of 8.1 MBq of ¹²³I or 111–370 MBq of ¹³¹I, with a subsequent ablative dosage of ¹³¹I ranging from 3.7 to 7.4 MBq (11). Using the 1-y follow-up scan, without a tg level, these workers reported that total ablation was obtained in 75% of patients receiving ¹²³I before the ablative dosage and 60% in those whose diagnostic tracer was ¹³¹I (P = 0.27). However, there is no indication what the mean ablative dosages were for these 2 groups and whether they were statistically different from each other. Park et al. did not observe stunning after they reduced the ¹³¹I activity administered to 74–111 MBq (11). ¹²³I used in the low dosages that Park et al. and our group administer provides excellent scintigraphic correlation with the ¹³¹I scintigraphs performed 7–10 d after ablation.

Several studies have found no evidence of stunning on ablation rate with ¹³¹I diagnostic activities of up to 185 MBq. The issue of stunning, and its clinical impact, may be related to the dosage of ¹³¹I used. Park et al. noted that qualitatively diagnosed stunning was not a problem at activities of 74–111 MBq (11).

Leger et al. imaged 51 patients after injection with 10–20 MBq (mean, 15.2 MBq) of ¹²³I, 19 of whom also received 185 MBq of ¹³¹I on the same day, yielding 93% and 94% ablation rates, respectively, using only scintigraphic evidence (12). Although stunning was documented only qualitatively in the group receiving diagnostic ¹³¹I, there was no effect on the outcome (12). Dam et al. used an average therapeutic dosage of ¹³¹I of 4.58 ± 0.21 GBq after a diagnostic dosage of 185 MBq, with visual evaluation of decreased uptake (stunning) detected in 19% of patients 7 d after therapy (13). The follow-up ablation rate was 88% in patients whose scans showed no evidence of stunning and 91% (P > 0.05) in patients with stunning. However, 26% of patients were lost to follow-up, so information on the therapeutic outcome was available only in 74% (13). Karam et al. also found no difference in ablative outcome, whether a diagnostic dose of 92.5 or 185 MBq of ¹³¹I was used, with ablation in 78% and 74% of patients, respectively (P > 0.05), although the tg assay was not used (14). Similarly, Morris et al. noted no difference in the success rate of ablation after diagnostic ¹³¹I dosages between 111 and 185 MBq given before therapeutic activities of 3.7–7.4 GBq of ¹³¹I (65% ablated) as compared with a group that had no diagnostic scans and, thus, no exposure to potentially stunning ¹³¹I before treatment (67% ablated) (15).

In contrast to these 4 articles, 2 studies suggest a stunning effect of ¹³¹I, in activities of 111–185 MBq, on the outcome of ablation, but both have flaws that raise questions about the authors' conclusions. Muratet et al. retrospectively reviewed 2 patient groups given diagnostic scanning dosages of either 37 or 111 MBq of ¹³¹I, which were then treated with 3.7 GBq of ¹³¹I (16). Follow-up studies performed 6.1–16.8 mo later revealed successful ablation in 50% of patients who had been scanned with 111 MBq of ¹³¹I and 76% success in the group receiving 37 MBq (P < 0.001). However, the group receiving the lower dosage was not matched for sex and contained a significantly higher percentage of men (16). In the second such study suggesting an effect of hypothesized stunning on outcome, Lees et al. retrospectively analyzed 3 patient groups that received, before remnant ablation with 3.7 GBq of ¹³¹I, a diagnostic scan dosage of 185 MBq of ¹³¹I, a diagnostic dosage of 740 MBq of ¹²³I, or no diagnostic radioiodine. Visual evidence of complete ablation was present in only 47% of patients receiving ¹³¹I, 86% of those receiving ¹²³I, and 83% of the group with no pretherapy exposure to a diagnostic dose of radioiodine. The 3 groups, however, had significant differences between them. The patients who received ¹³¹I for both diagnosis and therapy received a statistically larger amount of ¹³¹I therapy. The group with no pretherapy diagnostic scan had a mean serum tg of 10.7 ng/mL, whereas the first 2 groups had a far greater tumor burden with significantly higher tg values of 474 and 480 ng/mL, respectively. Follow-up after ablation imaging was performed as early as 3 mo after therapy when cell killing may not have been complete, so scintigraphy at this early time might lead falsely to the assumption of persistent (but mitotically sterile) tumor (17).

Finally, there is a study with the remarkable conclusion that the patients who experienced apparent thyroidal stunning had a better outcome than those who did not. Bajen et al. found that in 21% of patients studied, the posttherapy scan showed less uptake (stunning) qualitatively than on the diagnostic scans (18). A serum tg level of <3.0 ng/mL was used in this study, along with the radioiodine scan, to determine completeness of ablation. Sixty-two percent of stunned glands and only 37% of nonstunned glands were ablated according to these criteria. In this study, the therapeutic ¹³¹I dosage—ranging more widely than in any of the other studies (1.85–7.4 GBq) because patients with distant metastatic disease were included—was given an average of 7.2 wk after the diagnostic activity (185 MBq of ¹³¹I). This represents a far greater delay than in any of the other studies, wherein the time between diagnostic and therapeutic dosages of ¹³¹I was 1–9 d, except for that of Leger et al., which averaged 34 d with no evidence of an effect of stunning on ablation noted (12). Therefore, it is possible that enough time had passed for a therapeutic effect of the diagnostic activity to be added to the dosage actually given for treatment. The major problem in making any conclusions from the study of Bajen et al. arises from the fact that 23% and 26%, respectively, of the 2 groups had not yet had a follow-up scan (18).

There are also reports of diagnostic dosages of ¹²³I with activities of 185–200 MBq causing the stunning phenomenon (19,20), presumably from Auger electrons emitted by ¹²³I (21). However, despite the stunning phenomenon observed with both radiopharmaceuticals in a range of dosages, complete ablation of postthyroidectomy remnants can usually be successfully achieved.

We chose not to quantitate the uptake of therapeutic ¹³¹I given after diagnostic dosages of ¹²³I or ¹³¹I because this has been reproducibly done so often previously. Many practices now use recombinant human TSH (rhTSH) routinely for thyroid remnant ablation, but this study began before local insurance companies made rhTSH available to all patients. However, the lack of rhTSH usage should not have altered the results, because scintigraphy after thyroid hormone withdrawal or after rhTSH stimulation yields equivalent results (22).

Although ¹²³I is frequently used to image patients with thyroid cancer before ablation, ¹³¹I remains an important diagnostic radionuclide if one must perform dosimetric studies to determine the largest safe dose that can be given or currently when rhTSH is used. Furthermore, ¹²³I may not always be available in every area performing nuclear medicine, and thus we feel it is important to know an activity of ¹³¹I that may be given for diagnostic purposes without sequelae that could affect the therapeutic outcome.

CONCLUSION

This study has demonstrated that a preablative scanning dosage of 74 MBq of ¹³¹I yields the same thyroid remnant ablation rate as 14.8 MBq of ¹²³I, confirmed not only by the follow-up scan at 6–8 mo but also by a simultaneous undetectable level of serum tg. This is the first prospective study, to the author's knowledge, of the clinical import of radioiodine-induced stunning with the ¹²³I and ¹³¹I groups matched for age, sex ratio, tumor type and stage, and ablation dosage, and with virtually complete (98%) patient follow-up. If one must use ¹³¹I for preablative scanning, 74 MBq appears to be an appropriate activity that does not lower the ablation rate as compared with ¹²³I.