TO THE EDITOR:

In a recent article, de Keizer et al. demonstrated the radiation safety of radioiodine therapy with 7.4 GBq of ¹³¹I after stimulation with recombinant human thyroid-stimulating hormone (rhTSH) in thyroid cancer patients (1). For 14 consecutive patients (17 treatments), they found that the red marrow dose was 0.16 ± 0.07 mGy/MBq, thus delivering a maximum total radiation dose of 1.91 Gy to the red marrow. The mean blood dose was 1.69 ± 0.34 Gy. In 4 of 17 treatments, the calculated blood dose exceeded the limit of 2 Gy. The authors did not observe a case of hematologic toxicity.

In a recent international multicenter study (2), we assessed the absorbed blood dose to 9 patients with differentiated thyroid cancer who twice received a tracer dose of 74 MBq of ¹³¹I before ablation therapy. The blood dose was calculated as a sum of contributions using the measured residence times for the blood, remnant, and remainder of the body coupled with the appropriate S values. The blood residence time was determined through direct, sequential sampling, whereas the remnant and remainder residence times were obtained through sequential image analysis. The blood dose values for euthyroid patients after application of rhTSH were 0.08 ± 0.03 and 0.11 ± 0.03 mGy/MBq, assuming vessel radii of 0.02 and 0.5 cm, respectively (2).

Postulating that the iodine kinetics of 74 MBq of ¹³¹I are equal to those of 7.4 GBq of ¹³¹I, the expected blood dose range for 7.4 GBq of ¹³¹I is between 0.60 ± 0.21 and 0.82 ± 0.22 Gy. The mean value of 1.69 Gy reported by de Keizer et al. (1) is approximately 2–3 times higher.

We believe that some aspects degrade the reliability of the blood and red marrow dose values published by de Keizer (1). The first of these aspects is that the median total body residence time (RT_TB) of 155 h (value recalculated from Table 2, column 1, not 132 h as stated in the text [fourth paragraph of “Results”]) would be high even for healthy subjects or in radioiodine therapy of benign thyroid disease. According to the International Commission on Radiological Protection (3), the RT_TB is expected to be an order of magnitude lower in athyreotic patients. An overestimation of the RT_TB results in blood dose values that are too high. The high value for RT_TB reported by de Keizer et al. cannot be explained by the residence times in residual thyroid tissue or neoplastic cells that were published recently (4). For none of the patients did the uptake in thyroactive tissue exceed 4%, and the median effective half-life was 2.7 d. The mean effective half-life in the remainder of the body must have been even longer (∼4.5 d) to achieve an RT_TB of 155 h. Evaluations of residence time from direct measurements of whole-body activity over several days usually provide reliable data that are consistent with theory (e.g., in one study (2), the median RT_TB in euthyroid patients was 19 h). In the study of de Keizer, whole-body scintigraphy was performed at 24, 48, 120, 216, and 336 h after injection of the radioiodine (4). Although the 24-h scans might suffer from high-counting-rate dead-time effects and the accuracy of the extrapolation of the time-activity function to time zero might be degraded, it would be interesting to use the scans to recheck the RT_TB values given in (1).

A second aspect degrading the reliability of the values is that the authors do not list blood residence times. For ¹³¹I therapy of thyroid carcinoma, the blood residence time correlates strongly with RT_TB (2,5). Knowledge of the cumulated activity in blood would enable the reader to evaluate the reliability of the RT_TB and blood dose values.

A third aspect is that the blood activity was fitted with a long decay component (biologic half-life, 80 d) for the second term of a biexponential decay function. The procedure is adequate for healthy subjects but, in conjunction with a short sampling period of not more than 72 h, might overestimate the blood residence time in athyreotic patients and patients with small amounts of residual thyroactive tissue with a mean effective half-life of 2.7 d. It would have been more appropriate to calculate a range of residence times using a long decay constant and continuing the last measured decay rate to infinity. The actual value is expected to lie within this range, and this approach would allow the reader to assess the uncertainty of the results.

A fourth aspect is that de Keizer et al. evaluated the red marrow dose using a method described by Sgouros (6) for bone marrow dosimetry in radioimmunotherapy. Sgouros stated stringent conditions under which the equations in his publication are valid. The authors have not demonstrated the validity of the theory for radioiodine. Because the biokinetics of radioiodine (sodium iodide) are not comparable to those of the much larger molecules used in radioimmunotherapy, the red marrow dose values might be substantially incorrect.

In summary, the dose values for red marrow and blood have to be reevaluated because they indicate that the commonly accepted blood dose limit of 2 Gy might be reached or even exceeded. The outcome is contrary to the expectation of a reduced blood dose due to higher renal clearance and reduced effective half-life in euthyroid patients. In the future, instead of giving uncertain bone marrow dose estimates, we should determine blood doses as accurately as possible to see if the limits of the traditional radioiodine therapy guidelines (i.e., a maximum blood dose of 2 Gy) can be further extended when using rhTSH.