Comparison of thermal ablation and routine surgery for the treatment of papillary thyroid microcarcinoma: a systematic review and Meta-analysis

Abstract

Background

Thermal ablation (TA), as an alternative to surgery, has shown some benefits in the treatment of papillary thyroid microcarcinoma (PTMC) patients, especially for those who are at high risk for surgery or refuse surgery. We performed a systematic review and meta-analysis to evaluate the efficiency, safety, and economy of TA, compared with those of routine surgery (RS), for the treatment of PTMC.

Methods

PubMed, Embase, Cochrane Library, China National Knowledge Infrastructure (CNKI), Wanfang, and VIP databases were retrieved from inception to 10 January 2020 to identify relevant original studies on comparison of TA and RS for treatment of PTMC. The recurrence rate, recurrence-free survival (RFS), complication rate, operation time, postoperative length of stay, and cost during the perioperative period were extracted as main indices. The pooled standardized mean difference (SMD) or odds ratio (OR) with 95% confidence intervals (CI) were calculated and analyzed. Chi-square test and I² statistic were applied to determine the heterogeneity among studies. The sensitivity analysis was applied to explore the origin of heterogeneity, and the publication bias was evaluated by Egger’s test.

Results

Seven retrospective studies with a total of 867 patients met the eligibility criteria and were included in the final meta-analysis. Our study demonstrated that TA showed significant reduction in complication with a pooled OR 0.24 (95% CI 0.13 to 0.43), postoperative length of stay with a pooled SMD −3.14 (95% CI −4.77 to −1.51) and cost during the perioperative period with a pooled SMD of −1.69 (95% CI −3.18 to −0.20). It also demonstrated that both TA and RS had similar pooled proportion of recurrence of OR 0.93 (95% CI 0.38 to 2.30) and recurrence-free survive (RFS). The sensitivity analysis showed that each included study had no significant effect on the results and the results were stable and reliable. The Egger’s test demonstrated publication bias was acceptable.

Conclusions

TA may not be oncologically inferior to RS, and it is a relatively safe and economical alternative for the treatment of PTMC.

Keywords:

• Papillary thyroid microcarcinoma
• radiofrequency ablation
• laser ablation
• microwave ablation
• surgery
• meta-analysis

Introduction

Papillary thyroid microcarcinoma (PTMC), defined as 10 mm or less in maximum diameter, is a very common subtype of thyroid cancer with a favorable prognosis in general [1–3]. In recent years, due to the development of imaging technology, the number of incidentally discovered PTMC has increased rapidly, which sometimes causes the over-diagnosis or even over-treatment of this kind of indolent and stable cancer [4,5]. The American Thyroid Association (ATA) guidelines suggested that an active surveillance (AS) can be considered in PTMC patients without evident metastases or local spread, and in patients who have short remaining life span, high surgical risk [6]. However, local invasion or nodal metastases occurred in some patients with long-time AS [7]. Besides, the tumor could cause anxiety in patients when it is left untreated, which always makes patients trigger intensive treatment. For patients with PTMC, routine surgery (RS), such as thyroid lobectomy and total thyroidectomy, is the standard treatment, followed by thyroid hormone therapy and/or radioactive iodine therapy [8]. Nevertheless, the risk of complications associated with RS is high, particularly in patients with repeated neck dissections, and the cost of the RS is also remarkable [9]. Thus, thermal ablation (TA) techniques, including radiofrequency ablation (RFA) [10,11], microwave ablation (MWA) [12,13], and laser ablation (LA) [14–16] are suggested as alternatives for PTMC patients at a high risk of complications from RS or reject to undergo repeated surgeries for fear. A previous meta-analysis indicated that the therapeutic success rate of RFA for locally recurrent thyroid cancer is 100%, and serum thyroglobulin was reduced by 71.6% [17]. At the same time, these kinds of imaging-guided techniques also represent less invasiveness and unavoidable complications than the surgery, which counterbalances the situation of over-treatment [5].

Although TA has been shown as a favorable therapy modality for PTMC, several complications after the TA have been reported in recent years, including transient or persistent hoarseness, hypocalcemia, hematoma formation, and skin burns [17,18]. In addition, previous studies are not convincing for their short follow-up period, small sample size, and retrospective design. Therefore, it is necessary and timely to collect published clinical data regarding the use of TA for PTMC.

To the best of our knowledge, our present systematic review and meta-analysis is the first study, directly compared with the RS, to assess the efficiency, safety, and economy of TA for the treatment of PTMC. We evaluated recurrence rate, recurrence-free survival (RFS), complication rate, operation time, postoperative length of stay, and cost during the perioperative period with TA for the treatment of PTMC and compared the results with those of RS.

Materials and methods

Literature search strategy

A computerized search of 6 databases, namely PubMed, Embase, Cochrane Library, China National Knowledge Infrastructure (CNKI), Wanfang and VIP databases, was performed to identify relevant original studies on comparison of TA and RS for the treatment of PTMC from inception to 10 January 2020, without any language restrictions. The complete search strategy used for PubMed was: ‘(thyroid cancer [Title/Abstract] OR thyroid carcinoma [Title/Abstract] OR thyroid malignancy [Title/Abstract]) AND (ablation [Title/Abstract] OR ablation technique [Title/Abstract]) AND (surgery [Title/Abstract] OR operations [Title/Abstract])’. To identify other suitable articles, we also performed a manual search (including unpublished and gray literature) using the reference lists of key articles published.

Eligibility criteria

Studies according with all of the following criteria were included: (a) participants: patients with primary PTMC without clinically lymph node or distant metastasis; 10 mm or less in maximum diameter; no lymph node or distant metastasis; three or fewer tumors confirmed by ultrasonography (US); (b) intervention and control: the ablation method applied in the experimental group was RFA, MWA, or LA；the control group underwent RS (including thyroid lobectomy and total thyroidectomy) without any ablation intervention; (c) outcomes: the study compared TA with RS for the treatment of PTMC or demonstrated the clinical value of TA for PTMC; (d) studies: both prospective studies and retrospective studies were included; case reports, comments, editorials, letters, review articles, conference abstracts were excluded; there was no language restriction.

Study selection

Searches of the six databases and additional sources led to the identification of 5036 potentially relevant studies. Two experienced reviewers evaluated whether the titles and abstracts fulfilled the eligibility criteria. A full-text version was obtained if the information was insufficient to assess the exclusion or inclusion of an article. Final decisions on inclusion were made after browsing of the full text. Disagreements were solved by the discussion with a third reviewer. If sufficient information or data were unavailable in the potentially included studies, we would try to contact the original authors to obtain additional information, if possible.

Data extraction

The same two reviewers extracted the following data from the included studies independently, using a standard data extraction form: [1] study characteristics: the first author’s name, publication year, city, study design, and sample size; [2] demographic and clinical characteristics of the patients: mean age, tumor diameter, type of TA, and type of RS; [3] postoperative characteristics of the patients: the postoperative length of stay, cost during the perioperative period, complication rate, operation time, recurrence rate and RFS; [4] postoperative complications. The basic characteristics of each included article were shown in Table 1.

Table 1. Basic characteristics of each included article.

First author	Year	City^a	Study design	No. of patients (TA/RS)	Mean age (year) (TA/RS)	Mean tumor diameter (mm) (TA/RS)	Type of TA	Type of RS(TL/TT)^c	Complications	Assessment criteria^d	NOS^e
											Patient selection	Comparability	Outcome assessments	Sum of score
Cao	2018	Shanghai	RC^b	19/36	47.37 ± 11.95/47.53 ± 11.87	5.35 ± 1.57/7.70 ± 5.33	MWA	36/0	Voice change	①,②,③	***	**	**	7
Chen	2018	Hangzhou	RC^b	49/40	44.88 ± 11.04/45.78 ± 13.74	10 m or less in maximum diameter	MWA	40/0	Neck pain	①,②,③,⑥	***	**	***	8
Ma	2018	Hangzhou	RC^b	30/35	46.10 ± 8.20/49.97 ± 10.30	10 mm or less in maximum diameter	MWA	35/0	Voice change	③,⑥	***	**	**	7
Li	2018	Beijing	RC^b	46/46	43.63 ± 9.27/49.59 ± 9.00	4.49 ± 1.55/4.29 ± 1.37	MWA	37/9	Voice change	①,②,③,④	***	**	***	8
Li	2019	Beijing	RC^b	168/143	47.36 ± 10.75/49.18 ± 11.41	10 mm or less in maximum diameter	MWA	143/0	Voice change	③,④,⑤(5 years)	***	**	***	8
Zhang	2020	Beijing	RC^b	94/80	45.40 ± 10.80/44.10 ± 9.60	6.14 ± 2.54/6.03 ± 1.53	RFA	58/22	None	①,②,③,④,⑥	***	**	***	8
Zhou	2019	Shanghai	RC^b	36/45	41.50 ± 11.30/41.60 ± 9.60	4.70 ± 1.40/5.00 ± 1.10	LA	45/0	Hyperthyroidism	①,③,④,⑤(1,3 years),⑥	***	**	***	8

N/A: not available.

* represents one score in this area, **represents two scores in this area, and ***represents three scores in this area.

^a All are Chinese cities.

^b Retrospective cohort.

^c TL: Thyroid lobectomy; TT: Total thyroidectomy.

^d ①Postoperative length of stay; ②Cost during the perioperative period; ③Complication rate; ④Recurrence rate; ⑤Recurrence-free survival; ⑥Operation time.

^e Newcastle-Ottawa Scale.

Quality assessment

Two study reviewers assessed the methodological quality of the included studies independently, using the Newcastle–Ottawa Scale (NOS) [19]. Disagreements were also settled by consulting with the third reviewer. The result of the quality assessment was also shown in Table 1.

Data synthesis

We estimated standardized mean difference (SMD) with 95% confidence intervals (CI) or odds ratio (OR) with 95% CI for continuous variable and dichotomous variable, respectively. STATA version 14.0 (STATA Corporation, College Station, TX, USA) software was used to perform the statistical analysis. The recurrence rate, RFS, complication rate, opration time, postoperative length of stay, and cost during the perioperative period were extracted as the main outcomes for this systematic review and meta-analysis. For all analyses, p < 0.05 was considered statistically significant. Heterogeneity among studies was determined by the chi-square test for pooled estimates (p < 0.05, indicating significant heterogeneity) and the calculation of the I² statistic (0–40%, may have insignificant heterogeneity; 30–60%, 50–90%, and 75–100%, may have moderate, substantial and considerable heterogeneity, respectively) [20]. We adapted a random-effects model (DerSimonian-Laird method) when I² > 50% and p < 0.1, which indicated a significant heterogeneity across studies, otherwise a fixed-effects model (Mantel-Haenszel method) when I² < 50% and p > 0.1, which meant an absence of significant heterogeneity. Sensitivity analysis was applied to explore the origin of heterogeneity when significant heterogeneity was confirmed. If an outcome included more than five studies, publication bias was evaluated using Egger’s test.

Results

Characteristics of the included studies

Literature search

Seven eligible studies with a total sample size of 867 patients were included in the final systematic review and meta-analysis. Search and selection process were presented in a detailed flow chart (Figure 1). All of these original articles were retrospective studies, and the NOS quality was moderate overall. Regarding the demographic characteristics of the patients in the seven studies, the mean age ranged between 31 years and 60 years. The mean initial tumor diameter ranged 1 and 10 mm in 4 studies [21–24].The remaining three studies [25–27] didn’t list in detail, but had mentioned a maximum diameter of less than 10 mm. All studies reported complication rate after the TA or RS [21–27]. Four, two, four, and five studies reported the recurrence rate [22–24,27], RFS [24,27], cost during the perioperative period [21–23,25] and the postoperative length of stay [21–25], respectively. The characteristics and quality of the seven included studies were summarized in Table 1.

Figure 1. Procedure for the selection process.

Outcomes

Comparison of recurrence rate

Four studies with a total of 658 patients compared the recurrence rate. Two studies, from the same first author Li, reported patients’ recurrence rate in 42 months after the treatment while Zhang and Zhou reported in 60 and 36 months, respectively. Both TA and RS groups in Li’s first study reported a recurrence rate of 0(22). We collected the data from the four studies for analysis. The significant heterogeneity was nonexistent (p = 0.98, I² = 0%), so we adapted a fixed-effects model. The result showed that patients’ recurrence rate was similar under the two interventions with no statistically significant difference (OR 0.93, 95% CI 0.38 to 2.30, p = 0.88; Table 2, Figure 2).

Figure 2. Forest plot of the recurrence rate.

Table 2. Summary estimates of the meta-analytic pooled proportions.

	p-value for heterogeneity	I2	SMD or OR and pooled proportions [95% CI]	p-value for statistically significant difference
Recurrence rate	0.98	0%	0.93 [0.38, 2.30]	0.88
Complication rate	0.47	0%	0.24 [0.13, 0.43]	<0.001
Operation time	0.08	61%	−4.49 [−5.15, −3.83]	<0.001
Postoperative length of stay	<0.001	96%	−3.14 [−4.77, −1.51]	<0.001
Cost during the perioperative period	<0.001	97%	−1.69 [−3.18, −0.20]	0.03

Comparison of RFS

Two studies reported the RFS at different time points (Li’s in 5 years [27] and Zhou’s in 1 and 3 years [24]). In Li’s study, the 5-year RFS between MWA group (95.8%) and RS group (95.5%) showed no statistically significant difference (5-year: p = 0.97). In Zhou’s study, the 1- and 3-year RFS were 100% and 97.2% for LA group, respectively, and the corresponding values for the RS group were 100% and 93.3%, respectively. Both of the differences were not statistically significant (1-year: p = 1.00; 3-year: p = 0.82).

Comparison of complication rate

All studies reported the complication rate and the Egger’s funnel plot was generated to evaluate publication bias with a favorable result (p = 0.810; Figure 3(a)) In Zhang’s study, 3 patients reported complications in the RS group and none in the TA group [23]. A fixed-effects model was adapted due to the nonexistent significant heterogeneity (p = 0.47, I² = 0%). The analysis demonstrated that patients undergoing TA had a lower risk of complication after the treatment than the patients undergoing RS (OR 0.24, 95% CI 0.13 to 0.43, p < 0.001; Table 2, Figure 5(a)).

Figure 3. Egger’s funnel plot of (a) the complication rate; (b) the postoperative length of stay.

Comparison of operation time

Four studies with 409 patients reported the operation time. In Ma’s study, the operation time was not listed in detail, instead of which, they reported that the operation time of TA group ranged from 10 to 30 min while RS group was more than 30 min [26]. Thus, the remaining three studies were included in the meta-analysis, and the heterogeneity was significant (p = 0.08, I² = 61%). The sensitivity analysis was applied to detect the effects of each included study by repeating the meta-analysis while dislodging one study at a time. As shown in Figure 8(a), no included study had influenced the overall result, which could be considered to be stable and reliable. The random-effects model was adapted, and the difference of the overall effect was also significant (SMD −4.49, 95% CI −5.15 to −3.83, p < 0.001; Table 2, Figure 6), which meant that the operation time of TA group was significant shorter than that of RS group.

Figure 4. Forest plot the complication rate.

Comparison of postoperative length of stay

Five studies with a total of 491 patients reported the postoperative length of stay. In Zhang’s study, RFA was performed in an out-patient interventional US operating room without hospitalization while in the RS group, patients took 9.35 ± 4.39 days in the hospital [23]. The result showed a noted significant heterogeneity (p < 0.001, I² = 96%), so a random-effects model was adapted. The result of sensitivity analysis was also stable and reliable (Figure 5(b)). The result of Egger’s test showed an acceptable publication bias (p = 0.128; Figure 3(b)). Our analysis demonstrated that patients undergoing TA stayed shortly in the hospital after the treatment with a statistically significant difference (SMD −3.14, 95% CI −4.77 to −1.51, p < 0.001; Table 2, Figure 7).

Comparison of cost during the perioperative period

Four studies with a total of 410 patients discussed the cost during the perioperative period, and we converted all expenses into RMB according to the exchange rate. The considerable heterogeneity was also noted (p < 0.001, I² = 97%), thus a random-effects model was used. The sensitivity analysis was conducted with a reliable and stable result (Figure 5(c)). Our analysis showed that patients undergoing TA paid less for the treatment significantly compared with the RS group (SMD −1.69, 95% CI −3.18 to −0.20, p = 0.03; Table 2, Figure 8).

Discussion

In 2015, ATA guidelines recommended that very low risk PTMC, defined as a PTMC without local invasion or clinically evident metastases and no convincing cytologic evidence of an aggressive variant, could be followed with AS simply. In addition, the biopsy for sub-centimeter nodules, which are highly suspicious for PTMC on ultrasound, is not recommended to conduct to avoid the over-diagnosing PTMC according to the ATA guidelines [6]. However, there are no clinical, imaging, or biomarker that can accurately distinguish the small percentage of aggressive PTMC from the majority of indolent tumors [23]. As a result, a small percentage of PTMC patients’ tumor may ultimately develop enlargement, even extra-thyroidal extension [28,29], and distant metastasis [30]. Besides, PTMC patients are easy to feel anxious about tumor progression if no interventions were applied. For these reasons, a significant number of PTMC patients come under consideration for interventions.

TA, the local application of extremely high temperatures to cause irreversible cell injury, tumor apoptosis, and coagulative necrosis ultimately [31], has been a promising therapeutic modality for tumors. Particularly for small and/or unresectable tumors, TA has unique advantages, such as less invasiveness, lower complication rates, reduced morbidity, the possibility of performing the procedure on outpatients, and the suitability for real-time image guidance [32–35]. Our institution also conducts TA for the therapy of PTMC and finds its therapeutic effect (Figure 9).

Figure 5. Sensitivity analysis of (a) the operation time; (b) the postoperative length of stay; (c) the cost during the perioperative period.

Figure 6. Forest plot the operation time.

Figure 7. Forest plot the postoperative length of stay.

Figure 8. Forest plot the cost during the perioperative period.

In the new century, many original articles have explored the efficacy and safety of ablation for treating metastatic lymph nodes from PTMC [36–39]. Recently, two meta-analysis of single-arm tests also suggested that TA techniques have reasonable safety and efficacy profiles [40,41]. However, studies about the efficiency, safety, and economy of TA for PTMC are limited, and the effect directly compared with RS is worth assessing. Thus, we searched the six databases mentioned above and gray literature to obtain as much relevant research as possible. Seven retrospective studies with a total of 867 patients meet the eligibility criteria ultimately, and all of them were conducted in China.

In terms of safety, all the included studies [21–27] compared complication rate after the TA or RS and indicated that patients undergoing TA had a low complication rate with a pooled statistically significant difference (3.39% vs 12.94%, p < 0.001). The pooled proportion of complication was similar to a previous meta-analysis of single-arm tests (3.2% for TA) [40]. Voice change is the most common complication after TA and it may be caused by both thermal injury and hemorrhage or other factors such as inflammation, lidocaine injection, and fibrosis around the nerve [42,43]. Because of the particular position and the complexity of neck anatomy, RS performed on the thyroid is easy to cause the injury of recurrent laryngeal nerve. It is widely accepted that several methods, such as communication with patients, hydrodissection, sufficient knowledge of neck anatomy, necessary short-term pausesm, and moving shot technique may be helpful to avoid recurrent laryngeal nerve injury [44,45]. Hypothyroidism and hypocalcemia, caused during the treatment, are also common. The incidence of hypothyroidism, as reported, is as high as 75%, whereas that of permanent hypocalcemia ranges from 0% to 3.1% [46,47]. Thus, the 2017 Korea Thyroid Ablation Guideline [48] recommends ablation for PTMC to avoid complications after the treatment. The operation time is also a part of safety, because the longer operation time, in some ways, means more tissue damage, anesthetics and bleeding risk. Our analysis showed that the TA group had a significant shorter operation time than that of RS group (p < 0.001). The main reason was that TA could be performed in an outpatient clinic, and the procedures completed rapidly without general anesthesia.

Respectively, five [21–25] and four studies [21–23,25] compared the postoperative length of stay and the cost during the perioperative period with a pooled statistically significant difference: the TA group stay shortly in the hospital after the treatment and cost less, which indicates a faster postoperative recovery for PTMC patients. The cost of medical resources and patients’ treatment should be considered, particularly given PTMC’s high prevalence. Performing the TA procedures in an outpatient interventional US operating room may account for the difference in some ways [23,24]. Interestingly, we found that the postoperative length of stay of PTMC patients applying LA was shorter than those applying other kinds of TA. The reasons are as follows: First, the introducer needle (21-G) is smaller than the 16-G antenna used for MWA and 17-G needle for RFA, which means less invasive and is much easier to reach the target; In addition, the energy of LA is deployed predictably and precisely. The zone of necrosis obtained after LA is relatively small, which means more rapid recovery, shorter hospital stays, and lower cost ultimately. The significant heterogeneity was noted in the two outcomes (I² = 96%; I² = 97%), which may result from the different type of TA, imbalance of the different regional economic development, medical technology level, and operator experience in China. In Beijing and Shangai, the most medically advanced region of China, PTMC patients stayed shortly in the hospital and cost less during the treatment.

Regarding efficiency, four studies [22–24,27] measured the recurrence rate after the treatment at the time point from 36 to 60 months. The pooled recurrence rate was similar, and there was no statistically significant difference between the TA and RS group (2.91% vs 3.18%, p = 0.88) in our meta-analysis. TSH suppression therapy before the surgery and small central lymph node metastasis missing by the US before TA may account for the relatively lower recurrence in the RS group [23]. For patients with PTMC, recurrence is of greater concern than postoperative complications in most cases. And the US is the most frequently recommended imaging modality to detect the recurrence of thyroid cancer [49]. Several studies identified that some clinical characteristics, including gender, age, and tumor size, are responsible for recurrence of PTMC [50,51]. Besides, Takaki et al. speculated that there was an antitumor immune response induced by TA [52]. Some studies also suggested that TA could activate a so-called abscopal effect-spontaneous regression of a remote, non-targeted tumor [53,54]. Recurrent tumors found by the US after the initial treatment are successfully treated by repeat TA. In contrast, repeated surgery is usually tricky for recurrent tumors in the thyroid bed because of the distortion of normal tissue and fibrosis caused by scar tissue [27]. However, some scientists argued that TA may result in incomplete treatment compared with RS, thus the recurrence rates was higher and persistence or even distant metastasis would happen in long term follow-up [55]. Thus, a large sample and long follow-up time randomized controlled trial (RCT) is urgently needed to confirm the difference in recurrence rate between TA and RS.

Two studies (Li’s in 5 years [27] and Zhou’s in 1 and 3 years [24] measured the RFS, defined as no local recurrence or death during the follow-up period. Due to the limited number of the study, we conducted the systematic review. The differences of 1-, 3-, 5-year RFS between the TA group and RS group were not statistically significant (1-year: p = 1.00; 3-year: p = 0.82; 5-year: p = 0.97). The result indicated TA had no greater effect in RFS. It is also likely that oncologic outcomes will be similar for all interventions for PTMC at 5 years of follow-up, thus the need to minimize cost and complications is more considerable. Zhang’s study reported that 9 patients had occult PTMC which were not detected by US before surgery in the RS group, thus some occult PTMC may also be missed when performing RFA, which could make the RFS in the RFA group unfavorable [23]. In recent years, many studies based on the bioinformatics analysis were conducted to predict the RFS. Gu et al. reported that decreased EMX2OS (homeobox protein) expression might be a valuable prognostic biomarker of unfavorable RFS in classical PTMC [56], while Gong et al. found that preserved SCN4B might be a prognostic marker of favorable RFS in classical PTMC [57]. In general, the similar recurrence rate and RFS between the two groups in our systematic review and meta-analysis might be caused by the limited number of the study, small sample and insufficient follow-up time.

This meta-analysis also has some limitations of note. First, all the included studies were regionally concentrated in China. Therefore, the results of our meta-analysis may not be similar in other regions where medical technology and level may differ. If high quality studies in Europe, the United States, or any other countries could be conducted and included in this meta-analysis, the results would be steadier. Second, the number of included studies was relatively limited, and all of them were retrospective cohort. If RCTs could be involved, the overall effect of the meta-analysis will be more reliable. Third, factors that might affect the outcomes of treatment, such as operator experience and incision size, were not evaluated due to insufficient information or the difficulty for extracting accurate data. Forth, the heterogeneity in operation time, postoperative length of stay and cost during the perioperative period is relatively significant, which may be resulted from different regional economic levels, medical level difference, and age differences among included patients. Due to the limited number of the included studies in the three outcomes, subgroup analysis or meta regression could not be applied. If more similar studies could be carried out and included in the meta-analysis, sources of heterogeneity may be identified. Fifth, TA for PTMC is a relatively new technology, and the mean followed-up period is limited. Though our meta-analysis reported a low recurrence rate of TA, longer follow-up period studies are still needed to validate whether the results are durable or reproducible. Sixth, because of the novelty of ablation techniques, the differences of endpoints and definitions in the included articles do exist and may have an impact on the results. Thus, standardized terminology and data collection should be applied in this field, which enables a more reliable comparison among the different studies [58]. Seventh, for technical reasons, not all the gray literature, including conference abstracts, letters, and unpublished data, were retrieved, which may have caused a publication bias.

In conclusion, our meta-analysis is the first study, directly compared with RS, to explore efficiency, safety, and economy of TA for the treatment of PTMC. Our study demonstrates that TA, including both RFA, MWA, and LA, has a significant reduction in complication rate, operation time, postoperative length of stay, and cost during the perioperative period compared with RS. It also shows similar pooled proportions of recurrence and RFS with RS. Considering these results, TA may be an effective, relatively safe and economical treatment for PTMC patients, particularly in poor surgical contraindications or financial status.

Disclosure statement

The authors report no conflict of interest.

Figure 9. Ultrasound images of MWA in a patient with solitary PTC in our institution. (a) Before MWA therapy; (b) Immediately after MWA therapy; (c) 3 months after MWA therapy; (d) 18 months after MWA therapy.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [81902454].