Abstract
We read the article by Mujammami et al. with concern (1). The key principle of any treatment outcome study is based on the clarity of the patient selection criteria. If the treatment turns out to be promising, one must follow the same selection criteria in order to reproduce results. The post factum description of the patients who had received the treatment, that is, patients in the low- and intermediate-risk categories as defined by 2015 American Thyroid Association (ATA) classification (2), is not a reasonable substitute. The problem becomes clear when considering the intermediate-risk group (45% of all patients) in this report. The authors chose not to specify how many out of this group demonstrated no evidence of disease (NED) at follow-up. However, they report a pooled NED of 96.2%, while the published ranges for NED in low- and intermediate-risk are 86–91% and 52–63%, respectively. This discrepancy in NED with the expected ranges raises concern for unrevealed selection bias that led to misrepresentation of the intermediate-risk patients and inappropriate conclusion that 30 mCi ablation can be used for them.
It is important to recognize that 30 mCi activities are used specifically to treat non-cancerous remnant thyroid tissue (e.g., “remnant ablation”) (2). The 2015 ATA definition of “low-risk” patients would fit well with the remnant-ablation paradigm. However, if the risk of remnant tumor is negligible and patient survival is unaffected by ablation, why would anyone expose such patients to unnecessary radioiodine therapy? The current ATA guidelines do not recommend it routinely (2). In healthcare systems where a physician is financially rewarded for performance of a treatment procedure, there is a real danger that this study will be used as justification for the use of 131I treatment. This is especially likely given the fact that the authors do not even mention that there is no evidence-based proof that “ablation” has any survival benefit in the low-risk group, nor is it routinely recommended in the current ATA guidelines (2). While the authors might not have been able to predict 2015 ATA recommendations in 1998 when they started treating those patients, they were certainly aware of them at the time of the writing of the manuscript and chose not to discuss them.
About 45% of all patients in the study by Mujammami et al. were prepared for 131I ablation with recombinant human thyrotropin (rhTSH), which was locally available to patients who were intolerant to withdrawal or if withdrawal was contraindicated (3). It is important to clarify whether the patients in this study who received rhTSH indeed satisfied these criteria.
The authors used a “prospectively maintained registry” of patients from where they (retrospectively, of course) extracted cases. While “prospectively” may sound good, it is meaningless in the given context. The only meaningful use for “prospectively” is in the description of patient criteria-based selection for a trial. But since this registry exists, why not compare the outcomes of similar patients who received no 131I ablation and/or were treated with higher activities? That would be a truly scintillating work.