RESUMO
@#The use of high- or low-dose radio-iodine therapy (RAIT) for initial thyroid remnant ablation in post-thyroidectomised patients diagnosed with differentiated thyroid cancer (DTC) with no distant metastases has long been a subject of much debate. Meta-analyses and systematic reviews have been previously made using both randomised control trials (RCTs) and observational studies without due regard to differences in study design. Hence, amore focused meta-analysis of available RCTs alone was conducted to determine the presence of a compelling difference between the initial remnant ablation success rates of high- and low-dose RAIT in post-thyroidectomised DTC patient without distant demtastases. An extensive search of PubMed and Cochrane Central register of RCTs (up to August 2013) was performed by two reviewers, which was completed by hand search of referencesfrom releveangt articles and review papers published from 1996 to 2012. The two reviewers independtly selected eligible studies, with disagreement resolved by consensus. The inclusion criteria were as follows: (a) randomised controlled trials, (b) post-thyroidectomised adult subjects diagnosed with well differentiated thyroid cancer and no evidence of distant metastases, and (c) subject randomisation into 30-50 mCi or 100 mCi 131I treatment groups. Studies were exluded if (a) the full text of the study is not available, (b) the study is in another language other than English, and (c) if the data on relative risk was not available or could not be derived from the study. Of eight published RCTs on radio-iodine therapy as of August 2013, only 5 were eligible for this meta-analysis; namely those by JOhansen et al. (1991), Bal et al. (1996), Zaman et al. (2006), Maenpaa et al. (2008) and Caglar et al. (2012). The same two reviewers independenty extracted data from the full text of the selected five studies. Two-by-two tables comparing frequencies of successful and failed remnant ablation using low-dose (30-60 mCi) and high-dise (100 mCi) RAIT were derived from the published results of the included studies, and the weighted and pooled relative risks for successful remnant ablation were computed via the Mantel-Haenszel method using a fixed effects model (cx = 5%). Subgroup analyses were performed based on different definitions of a successful remnant ablation. The pooled relative risk (-0.03) was statistically insignificant (p=0.54) and had poor precision (95% confidence interval of {-0.12,0.06}) even when adjustments to the varied definitions of a successful ablation were performed. Thus, using available RCTs that compare high- and low-dose RAIT for remnant ablation of DTC, there is an apparent trend favoring higher success rates using high-dose RAIT. However, the lack of well designed RCTs precludes recommending high-dose initial RAI ablation, and encourages the present practice of individualized.