High versus low radioiodine activity in patients with differentiated thyroid cancer: A meta-analysis

Abstract

Background. The purpose of the meta-analysis was to estimate the effectiveness and toxicity of low activity radioiodine ablation versus high activity in patients with differentiated thyroid cancer (DTC). Design. A systematic review and meta-analysis was performed by including all randomized trials of low activity versus high activity radioiodine ablation after thyroidectomy. Standard meta-analytic procedures were used to analyze the study outcomes. Results. Ten trials were considered eligible and were further analyzed. The pooled risk ratio (RR) of having a successful ablation for an activity of 1100 MBq versus 3700 MBq (seven trials, 1772 patients) was 0.94 (95% CI 0.85–1.04, p-value = 0.21). The RR for successful ablation when only thyroid hormone withdrawal was used (five trials, 1116 patients) was 0.87 (95% CI 0.72–1.06, p-value = 0.17) and it was comparable to RR when only recombinant-human TSH (rec-hTSH) (two trials, 812 patients) was used (1.00, 95% CI 0.93–1.07, p-value = 0.92). Salivary dysfunction, nausea, and neck pain were significantly more frequent among patients with higher dose for ablation. Conclusion. Our meta-analysis provides some evidence from randomized trials that a lower activity of radioiodine ablation is as effective as higher dose after surgery in patients with DTC with lower toxicity.

Cancer of the thyroid is relatively uncommon, accounting for about 1% of all cancers worldwide [1]. However, during the past several decades, an increasing incidence of thyroid cancer has been reported in many parts of the world [2–4], a phenomenon which has been associated with more intensive diagnostic activities and change in radiation exposure in childhood and adolescence [5].

The vast majority of the cases are differentiated thyroid cancer (DTC), which is associated with a high 10-year survival rate (90–95%) [6]. The gold standard therapy for DTC is total or near-total thyroidectomy. Surgery is usually followed by the administration of radioiodine ablation aimed at ablating any remnant thyroid tissue, thereby improving the sensitivity of post-operative ¹³¹I scanning and serum thyroglobulin measurements for detection of recurrent disease and at eradicating potential microscopic residual tumor.

There is conflicting evidence about the role of radioiodine in reducing risk for recurrence in differentiated thyroid cancer [7]. However, this treatment is recommended from current guidelines [8–10] mainly due to the fact that the omission of radioiodine therapy could limit the sensitivity of post-treatment whole-body scanning and as a result the detection of persistent or metastatic disease [11].

Considering the unclear role of radioiodine in terms of recurrence, this intervention should at least offer an attractive risk-benefit ratio, namely low toxicity and no influence on quality of life. This is the reason why a reduced dose of radioiodine has been investigated. Indeed, lower dose of radioiodine has remarkable advantages, including lower toxicity and, consequently, better quality of life, less time in hospital isolation, as well as a lower radiation dose to extra-thyroidal compartments.

Several randomized trials investigating the concept of lower radioiodine activity for ablation have been published [12–17]. A meta-analysis including both observational and randomized studies could not come up with a reliable conclusion due to the low quality of the observational studies and the small number of patients in the randomized studies [18]. In fact, there is insufficient evidence to support a recommendation for or against a definite dose of ¹³¹I for ablation therapy. Recently, three large randomized studies have been published and presented more mature data on whether radioiodine ablation with lower activity is as effective as higher dose [19–21].

We, therefore, performed a systematic review and meta-analysis of all randomized evidence in order to estimate the effectiveness and toxicity of low activity radioiodine ablation versus high activity in patients with DTC.

Methods

Search strategy

Two independent investigators searched MEDLINE, and the Cochrane Controlled Trials Register, without language or year restriction through May 2012. The search strategy including the terms: thyroid cancer, radioiodine, and ablation.

In addition, the reference lists of all eligible studies as long as reference lists from prior relevant reviews were also scrutinized to identify relevant articles missed by the electronic searches.

Selection criteria

Only randomized controlled trials in which patients with differentiated thyroid cancer were allocated to a low activity versus a high activity radioiodine ablation after thyroidectomy were considered eligible. We defined as low activity the use of 1100 MBq and as high any activity higher than 1100 Mbq. We excluded non-randomized studies and randomized studies which have been presented only at meetings and have not been published as full-text. We also excluded randomized trials that used activities higher or lower than 1100 MBq as low activities.

Whenever multiple records were related to the same study, end point data was extracted from the report with the longest follow-up to avoid duplication of information in the meta-analysis calculations.

Data extraction

Data extraction was conducted independently by the two authors (AV and AN) and consensus was achieved to all extracted data.

From each eligible trial, we extracted the following data in a pre-specified database: author names, journal and year of publication, country of origin, inclusive dates of patient enrolment and number of centers involved. Additionally, we recorded the following items for both arms of each eligible trial: number of patients randomly assigned to treatment and analyzed per arm, age, T status, N status, type of thyroid cancer (papillary or follicular), type of surgery, definition of successful ablation, time from surgery to ablation, use of low-iodine diet, activity of radioiodine used, and method for thyrotropin (TSH) stimulation.

Risk of bias and publication bias assessment

Cochrane's risk of bias tool has been utilized in order to assess the individual risk of bias of each study [22]. The criteria used for quality assessment were sequence generation of allocation, allocation concealment, masking of participants, personnel, and outcome assessors, incomplete outcome data, selective outcome reporting, and other sources of bias. The two authors independently assessed the risk of bias in each eligible trial. We define an eligible trial as low quality trial if two or more criteria for quality assessment were categorized as high or unclear risk of bias.

Publication bias was assessed with the construction of contour enhanced funnel plots.

Outcome measures

The primary outcome of this study was to assess the successful ablation rate between low and high radioiodine dose, as defined by each eligible study.

Secondary outcomes were number of recurrences, and early adverse events (within one week from ablation) related to radioiodine ablation, including salivary dysfunction, neck pain, lacrimal dysfunction, nausea and any serious adverse events.

Statistical analysis

All the assessed outcomes were dichotomous variables. Results were expressed as risk ratios (RR) with 95% confidence intervals (CI). Analyses were by intention to treat and included all participants.

^χ2 test of heterogeneity and the I² statistic of inconsistency were used to assess heterogeneity between studies. Statistically significant heterogeneity was defined as a χ² p-value less than 0.1 or an I² statistic greater than 50%. In the absence of heterogeneity, pooled estimates of RRs with their 95% CIs were calculated using the Mantel-Haenszel method. In the presence of heterogeneity, the DerSimonian and Laird random effects method was used to pool primary studies estimates.

Pre-specified subgroup analysis was performed for primary outcome based on method for TSH stimulation [thyroid hormone withdrawal or use of recombinant human thyrotropin (rec-hTSH)]. A sensitivity analysis was performed by excluding the trials which accepted as surgical procedure not only the total or near total thyroidectomy but also the subtotal thyroidectomy. An additional sensitivity analysis was performed by including only trials with a modern definition of successful ablation according to guidelines used in current clinical practice namely an absence of visible uptake on a subsequent diagnostic whole-body scan and an undetectable stimulated serum thyroglobulin [23]. To assess the potential impact of quality of studies on outcomes, we performed an additional sensitivity analysis with the exclusion of low quality studies.

All statistical analyses were performed by using the Review Manager software (Version 5.0. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2008).

Results

Study selection and characteristics

The electronic search yielded 1341 reports. Of these, 1320 were excluded on the basis of the abstract or title leading to 21 potentially eligible trials. After scrutiny, eight randomized controlled trials [12–14,16,17,19–21] were considered eligible according to the inclusion criteria, whereas hand-searches in references of prior reviews on the same topic resulted in one additional trial [15]. In total, nine trials were considered eligible and further analyzed (Figure 1).

Figure 1. Flow chart diagram of study selection.

Baseline characteristics of patients included in the eligible trials are presented in Table I. Thyroid hormone withdrawal was chosen for TSH stimulation in six trials [12,14–17,20], while in three trials TSH stimulation was performed either by withdrawal or by rec-hTSH [13,20,21]. Two trials were randomized not only to high versus low radioiodine dose but also to two different methods for TSH stimulation (withdrawal or rec-hTSH) [20,21]. Three trials clearly stated that they recommended a low-iodine diet [17,19,20]. A variation of ablation activities were used and compared in eligible trials. Six trials compared the use of 1100 MBq as low activity versus 3700 MBq as high activity [12,13,17,19–21]. Two trials examined the success of ablation in relation to several administered activities ranged from 1100 to 5350 in one trial [14] and from 555 to 1850 in the other trial [16]. One trial used as low activity 1100 Mbq with comparison to 2220 MBq as high dose [15].

Table I. Characteristics of eligible trials.

Author [ref]	Year	Extent of thyroidectomy	Timing from operation to ablation	Timing from ablation to diagnostic scan	Method for TSH stimulation	Definition of successful ablation	Low-iodine diet	Total no. of patients	Ablation activities (MBq)
Creutzig [12]	1987	Total	3 weeks	3 months	Withdrawal	Neck uptake </ = 0.5%	NR	20	1100 vs. 3700
Johansen [13]	1991	Total and subtotal	1.5 month (median)	NR	Withdrawal and rec-hTSH	Negative WBS and thyroglobulin < 5 μg/l	No	63	1100 vs. 3700
Bal [14]	1996	Total and subtotal	NR	6–12 months	Withdrawal	Negative WBS, neck uptake < 0.2%, thyroglobulin < 10 ng/ml	NR	119	1100 vs. 1850 vs. 3330 vs. 5735 ^a
Gawkowska-Suwinska [15]	2001	Total	NR	6–8	Withdrawal	Negative neck scan and thyroglobulin < 4 ng/ml	NR	98	1100 vs. 2220
Bal [16]	2004	Total and subtotal	2 months (median)	6	Withdrawal	Negative WBS, neck uptake < 0.2%, thyroglobulin < 10 ng/ml^b	No	509	555 vs. 740 vs. 925 vs. 1100 vs. 1295 vs. 1480 vs. 1665 vs. 1850
Mäenpää [17]	2008	Total or near total	5–6 weeks	4–8 months	Withdrawal	Negative WBS and thyroglobulin < 1 ng/ml	Yes	158	1100 vs. 3700
Fallahi [19]	2012	Total or near total	4–6 weeks	6–7 months	Withdrawal	Negative WBS and thyroglobulin < 2 ng/ml	Yes	341	1100 vs. 3700
Mallick [20]	2012	Total and near total	1–6 months	6–9 months	Withdrawal and rec-hTSH	Negative WBS and thyroglobulin < 2 ng/ml	Yes	438	1100 vs. 3700
Schlumberger [21]	2012	Total	NR	8 ± 2 months	Withdrawal and rec-hTSH	Negative WBS and negative neck ultrasound and thyroglobulin < 1 ng/ml	NR	752	1100 vs. 3700

no, number; NR, not-reported; MBq, megabecquerel; rec-hTSH, recombinant human thyroid-stimulating hormone; TSH, thyroid-stimulating hormone; WBS, whole-body scan.

^a The ablation activities of 1100, 1850, 3300, and 5735 are the mean levels from four groups: 25–34, 35–64, 65–119, and 120–200 mCi; ^b Two of three criteria were required.

Four trials [17,19–21] defined the successful ablation according to modern guidelines that have been used in current clinical practice.

Primary outcome: Successful ablation rate

Due to the fact that the dose level that defines low or high differs greatly between eligible trials and in order to avoid the between-study heterogeneity, we made two separate analyses. The first analysis included trials that used 1100 MBq as low activity versus 3700 MBq as high activity and in second analysis trials used 1100 MBq versus 1850 MBq. We could not proceed to other dose comparisons due to the lack of data from more than one study in each dose comparison.

The pooled RR of having a successful ablation for an activity of 1100 MBq versus 3700 MBq (seven trials, 1772 patients) was 0.94 (95% CI 0.85–1.04, p-value = 0.21) (Figure 2A). The trial by Bal et al. [14] compared a low activity of 1100 MBq versus a high activity of 3330 MBq. This trial was not included in the first analysis because of the lower dose in the high-dose arm that used. The inclusion of this trial on the analysis did not alter the RR (0.93, 95% CI 0.85–1.03, p-value = 0.16).

Figure 2. Forest plot of risk ratio for successful ablation: 1100 MBq versus 3700 MBq. Two trials [21,22] were randomized to not only high versus low dose but also rec-hTSH versus withdrawal for TSH-stimulation. Schlumberger 2012a and Mallick 2012a presented data when rec-hTSH was used while Schlumberger 2012b and Mallick 2012b when withdrawal was used.

The pooled RR of successful ablation for an activity of 1100 MBq versus 1850 MBq (two trials, 249 patients) was 0.97 (95% CI 0.83–1.14, p-value = 0.73) (Figure 2B).

Subgroup analysis according to the method for TSH stimulation was performed for the 1100 MBq versus 3700 MBq analysis. The RR for successful ablation when only thyroid hormone withdrawal was used (five trials, 1116 patients) was 0.87 (95% CI 0.72–1.06, p-value = 0.17) and it was comparable to RR when only rec-hTSH (two trials, 812 patients) was used (1.00, 95% CI 0.93–1.07, p-value = 0.92).

The exclusion of trials with subtotal thyroidectomies did not alter the RR for successful ablation rate (five trials, 1709 patients) (0.93, 95% CI 0.83–1.03, p-value = 0.17). When the analysis was restricted only in the four trials with a modern and adequate successful ablation definition, the RR for successful ablation rate remained non-significant (0.93, 95% CI 0.83–1.04, p-value = 0.19).

Secondary outcomes

Three trials reported data on recurrences [17,19,22]. The trials had significant clinical heterogeneity regarding the median follow-up that ranged from 12 to 51 months. The RR for recurrence (1100 MBq vs. 3700 MBq) was 0.67 (95% CI 0.32–1.40, p-value = 0.29).

An overview of early adverse events related to ablation and the RR regarding the activity used (1100 MBq vs. 3700 MBq) is shown on Table II. Salivary dysfunction, nausea, and neck pain were significantly more frequent among patients with higher activity for ablation (p-values 0.03, < 0.001, and 0.003, respectively).

Table II. Meta-analysis of early adverse events related to radioiodine ablation.

Adverse events	1100 MBq (n/N)	3700 MBq (n/N)	No. of studies included	RR (95% CI)	p-value
Salivary dysfunction^a	56/653	79/649	3	0.71 (0.52–0.97)	0.03
Nausea	19/288	63/285	2	0.30 (0.19–0.48)	<0.001
Neck pain	36/288	62/285	2	0.57 (0.40–0.82)	0.003
Lacrimal dysfunction^b	58/433	72/431	2	0.81 (0.59–1.10)	0.18
SAEs^c	5/585	6/582	2	0.88 (0.05–16–72)	0.93

n, number of events; N, total number of patients; RR, risk ratio; SAEs, serious adverse events.

^aSalivary dysfunction also includes pain in salivary glands, sialadenitis or other salivary dysfunction; ^bLacrimal dysfunction includes dry eyes as well; ^cThe serious adverse events were not thought from the primary investigators to be related to trial´s interventions.

Risk of bias and publication bias assessment

We used Cochrane's risk of bias tool to assess the methodological quality of the trials (Figure 3). Among eligible trials, six [14,16,17,19–21] reported an adequate randomization mode, while only one [20] clearly reported blinding of outcome assessment. In addition, only one trial [19] reported that it was designed as a double-blind trial.

Figure 3. Forest plot of risk ratio for successful ablation: 1100 MBq versus 1850 MBq.

Only two trials were classified as high quality, according to the definition that described on the method section [19,20]. A sensitivity analysis including only these two trials showed a RR for successful ablation (for the comparison 1100 MBq vs. 3700 MBq) of 0.85 (95% CI 0.65–1.11, p-value = 0.23), thereby comparable to the RR when all eligible trials included at the analysis regardless of their quality.

In all meta-analyses funnel plots were symmetrical, indicating that publication bias was unlikely to have had a major influence in the analyses (plots not shown).

Discussion

Our meta-analysis summarizes all the available randomized evidence about the effectiveness of low activity radioiodine ablation versus high dose in patients with DTC. Our results, based on a large number of patients, indicate that the low activity (1100 MBq) has similar effectiveness in terms of ablation rate with high activity (3700 MBq), which is the current recommended dose [8,10], with better toxicity profile. A lower activity for radioiodine ablation offers some attractive advantages. First, it is associated with less time in hospital isolation, thus less cost for the healthcare system [17,20]. In addition, a lower dose limits the radiation exposure and this could have an effect not only in acute toxicity, as confirmed by our meta-analysis, but also in the development of second primary cancers [24,25]. The increased risk of secondary primary cancer in patients with DTC after ablation is a worrisome phenomenon considering the expected long-term survival in these patients and the fact that the ablation has not yet been proved to reduce risk for thyroid cancer recurrences [6].

As a result, a lower activity seems to be a justifiable option in patients with DTC. However, two key questions need to be addressed. The first question is whether radioiodine ablation should be offered as standard treatment strategy to all these patients. A recent observational study with more than 10-year follow-up found that the use of radioiodine did not prolong overall or disease-free survival in low-risk patients [26]. However, this question is difficult to be answered with existing data due to the lack of randomized evidence. In fact, there are only observational studies which are inherently more prone to bias [6,7]. Recently, a randomized trial has been designed in order to address the safety of avoiding post-operative radioiodine ablation in these patients [27]. The second question is whether a lower ablation activity can affect the recurrence rate. In our meta-analysis, only three trials presented data on recurrence and the RR for recurrence of low versus high activity was not statistical significant. However, a significant between-study clinical heterogeneity should be outlined since the follow-up ranged from 12 to 51 months. A follow-up of 12 months cannot be considered long enough to reveal any differences in recurrence rates. In the study with longer follow-up [17] no differences were observed between low and high activity in terms of regional or distant recurrences. In addition, observational studies with longer follow-up could not find any differences in disease-free survival in different radioiodine doses [28]. However, a firm answer to this question is expected from the well-designed randomized trials, which have been included in our meta-analysis, and are planning to present data on recurrences as secondary outcomes after longer follow-up [20,21].

The effectiveness of low and high activity ablation remained similar irrespectively the method which used to stimulate TSH, namely thyroid hormone withdrawal or rec-hTSH. The use of rec-hTSH has been associated with lower risk for hypothyroidism and preservation of quality of life [20,21,29,30]. However, further parameters (higher cost, need for more clinic visits) should be taking into account before a clear recommendation in favor of rec-hTSH can be made [31].

We have to acknowledge that several limitations exist in our meta-analysis. First, our analysis is based on published or presented data and not on individual patient data. Individual patient data meta-analysis is considered more reliable [32]. Second, eligible trials used different criteria to define successful ablation. However, when we restricted our meta-analysis on trials with a modern and adequate successful ablation definition, we could not find any difference between low and high activity. In addition, we were not able to make several comparisons among different dose activities due to the lack of data. This is the reason why we proceeded to two different comparisons: 1100 MBq versus 3700 MBq and 1100 MBq versus 1850 MBq. Finally, most of the eligible trials were categorized as low quality and one could argue that this could potentially influence the validity of the results. However, a sensitivity analysis including only high quality trials, according to our strict criteria, could not reveal any difference in successful ablation rate between low and high activity.

In conclusion, our meta-analysis provides some evidence from randomized trials that a lower activity (1100 MBq) of radioiodine ablation is as effective as higher activity after surgery in patients with DTC. These findings, if they confirm by further studies with longer follow-up, could have a great impact on clinical decision making for patients with DTC considering the benefits of using lower activity both for patients and healthcare providers. Future randomized trials should be concentrated on finding subgroups of patients in whom omission of ablation is a safe option.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.