Total Thyroidectomy Versus Partial Thyroidectomy for Non-Toxic Multinodular Goiter: A Systematic Review and Meta-Analysis of Randomized Controlled Trials

Abstract

Thyroid follicular nodular disease (non-toxic multinodular goiter), the commonest thyroid disorder, can be managed with different surgical methods, including total thyroidectomy (TT), near-total thyroidectomy (NTT), subtotal thyroidectomy (STT), and Dunhill operation (DO). This systematic review and meta-analysis aimed to provide an update on the role of TT versus NTT, STT, or DO in MNG concerning goiter recurrence, thyroid cancer incidence, and reported complications. We conducted a systematic literature search in Medline, EMBASE, and CENTRAL. We included randomized controlled trials (RCTs) that compared TT to partial thyroidectomy (NTT, STT, or DO) in the management of multinodular goiter (MNG). The following outcomes, goiter recurrence, cancer incidence, and adverse events (hypoparathyroidism and recurrent laryngeal nerve injury), were evaluated. We used the risk ratio (RR) to represent the dichotomous outcome. Subgroup analysis was performed based on the different types of partial thyroidectomy (NTT, STT, and DO). A total of 7 RCTs that included 1909 individuals were deemed eligible. TT showed significantly better results compared to PT in terms of goiter recurrence (RR = 0.05, 95% CI 0.02 to 0.13, P = 0.001; I² = 0%), whereas it showed similar results compared to PT in terms of thyroid cancer incidence (RR = 1.09, 95% CI 0.76 to 1.57, P = 0.63; I² = 0%). PT was significantly better than TT concerning transient adverse events (RR = 2.18, 95% CI 1.64 to 2.89, P = 0.001: I² = 63%). This meta-analysis showed that TT has a lower risk of goiter recurrence. TT is comparable to PT in terms of persistent adverse events but has a higher risk for transient adverse events.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13193-024-02057-y.

Introduction

Thyroid follicular nodular disease (FND), formerly known as non-toxic multinodular goiter or thyroid gland multinodular hyperplasia, is the most common thyroid disorder. It is a benign condition not associated with clinical changes in thyroid function [1, 2]. Thyroid enlargement due to follicular nodular disease can result in compressive symptoms, including dyspnea, hoarseness, and dysphagia. In most cases, surgical intervention may be required for compressive symptoms, suspicion of thyroid malignancy, cosmetic reasons, or progressive enlargement [1]. The surgical treatment reduces the goiter significantly and provides immediate symptom relief; however, it may pose the risk of developing some adverse events, such as hypoparathyroidism and recurrent laryngeal nerve palsy [1, 3]. Among the multiple surgical methods for the excision of follicular nodular disease (non-toxic multinodular goiter), including total thyroidectomy (TT), near-total thyroidectomy (NTT), subtotal thyroidectomy (STT), and hemithyroidectomy plus subtotal resection (Dunhill operation), the optimal surgical technique remains controversial [1, 3, 4]. Recent trials have demonstrated the superiority of NTT, STT, and Dunhill operation (DO) over TT in terms of avoiding temporary postoperative hypocalcemia [4–10]. Other clinical trials concluded that TT carries the advantage of being a one-step procedure that has a lower risk of goiter recurrence and avoids reoperation for the incidental finding of thyroid malignancy compared to STT and DO [9–11]. Despite the aforementioned advantages, recent systematic reviews of RCTs were conducted with conflicting results [12–15]. Moreover, the previous reviews were limited by their relatively low number of included studies and short follow-up time. Some RCTs have considered NTT as TT, whereas other studies considered it partial thyroidectomy (PT) [4, 5, 8]. To our knowledge, this is the first systematic review and meta-analysis to consider near-total thyroidectomy, subtotal thyroidectomy, and Dunhill operation as PT. Three RCT studies, representing 939 participants and comparing TT to PT surgical techniques, were introduced to the literature since the most recent systematic review [7, 11, 12].

This systematic review and meta-analysis aimed to provide an update on the role of TT compared to NTT, STT, and DO in patients with unilateral or bilateral follicular nodular disease (MNG) concerning goiter recurrence, thyroid cancer incidence, and reported complications.

Materials and Methods

Study Registration

This systematic review was carried out according to a pre-specified protocol and reported following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) checklist [16]. The pre-established protocol of this systematic review was registered in PROSPERO (CRD42022314751).

Eligibility Criteria

Participants

Individuals with unilateral or bilateral MNG received primary thyroidectomy; intervention: TT; Comparison: PT techniques (i.e., NTT, STT, and DO); prespecified outcomes: goiter recurrence, thyroid cancer incidence, and adverse events; and study design: RCTs. We excluded RCTs that have considered NTT as TT or compared one of the previously mentioned PT techniques with another PT.

Search Strategy

We searched Medline, EMBASE, and the Cochrane Central Register of Controlled Trials (CENTRAL) from the database inception to February 25, 2022, without any restriction on date. The complete search strategy is provided in the supplementary material.

Study Selection and Data Extraction

A total of four reviewers performed titles and abstracts screening against the eligibility criteria as well as full-text assessment and data extraction from the eligible trials. Every two reviewers work independently and in duplicate. Any disagreement was solved by discussion or a third reviewer's decision if consensus was not achieved. Reviewers extracted the following data: name of the first author and the year of publication, number of participants in each arm, gender, mean age, study population, mean follow-up period, treatment protocol, levothyroxine therapy, goiter recurrence diagnostic criteria used by each trial, and the desired outcomes reported by each trial.

Meta-Analysis

Data meta-analysis was performed using RevMan (Review Manager) version 5.4 (Cochrane Collaboration). All statistical analyses were performed using the random effects model. We used a confidence level of 95% with P < 0.05 for statistical significance. The statistical heterogeneity was assessed using I² and the P of the Chi² test (P ≤ 0.1 or I² test ≥ 50% indicates significant heterogeneity). The dichotomous outcome of goiter recurrence, cancer incidence, and adverse events was represented as risk ratio (RR) and pooled using the inverse variance weighting method. We performed subgroup analysis based on different types of PT techniques: NTT, ST, and DO. The quality of the evidence for each outcome was assessed using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) criteria approach and then classified as high, moderate, low, or very low quality.

Risk of Bias Assessment and Publication Bias

Reviewers performed the risk of bias assessment for the included RCTs using the Revised Cochrane Risk of Bias Assessment Tool to assess the risk of bias in eligible studies and classified studies into the following categories: high risk of bias, some concerns, or low risk of bias [17]. The potential of publication bias for each outcome was assessed by visual inspection of the funnel plot and symmetry assessment. Evidence of publication bias is considered present when it is asymmetrical. Furthermore, we manually searched the reference lists of the included RCTs for potentially relevant trials missed through the systematic search to avoid publication bias.

Results

The initial search yielded 5869 articles, of which 1876 duplicates were excluded, resulting in 3993 articles. A total of 3982 articles were further excluded after the title and abstract eligibility screening. Of 11 studies assessed for full-text assessment, four were excluded. This resulted in the final inclusion of seven trials in this systematic review and meta-analysis [4–6, 8–11]. Figure 1 shows the flow diagram for study selection. Details of the search strategy are provided in Supplementary Table A.

Fig. 1.

Study flow diagram

Trial Characteristics

A total of 1483 individuals were included in this systematic review. Of those, 649 (43.76%) were randomized to TT, and 834 (56.23%) were randomized to PT. The weighted mean age was 40.85 years (range 31.5–50.2) in the TT group and 41.22 (range 32.45–50) in the PT group. Most of the participants were female (76.53%). The mean follow-up time for most of the included studies was 5 years. The details of the treatment protocol specific for each trial are shown in Table 1.

Table 1.

Characteristics of the included studies

Study, year		Erbil 2006 [6]	Pappalardo 1998 [10]	Barcynski 2018 [12]	Mobayen 2015 [11]	Barcynski 2010 [9]	FAIZI 2015 [7]	Unalp 2009 [5]
Study arms		TT vs NTT	TT vs STT	TT vs DO and STT	TT vs DO	TT vs DO and STT	TT vs ST	TT vs NTT
Number of partcipants	Total thyroidectomy	104	69	177	90	191	120	75
	Partial thyroidectomy	NTT: 112	STT: 72	DO:174, BST: 175	DO: 83	DO: 189, BST: 190	ST: 120	NTT: 68
Gender	Male	TT: 10, NTT: 14	TT: 20, STT: 21	TT: 15, DO: 18, BST: 15	DO: 15, TT: 13	TT: 17, DO: 20, BST: 16	TT: 38, ST: 41	TT: 64. NTT: 59
	Female	TT: 94, NTT: 98	TT: 49, STT: 51	TT: 162, DO: 156, BST: 160	DO: 68 TT: 77	TT: 174, DO: 169, BST: 174	TT: 82, ST: 79	TT: 11, NTT: 9
Mean age (± SD)	Total thyroidectomy	47.3 (± 12)	48	TT: 45.9 (± 13.9)	43.1 (± 11.3)	46.51 (± 14.11)	31.52 (± 8.7)	50.2 (± 11)
	Partial thyroidectomy	45.7 (± 11)	50	DO: 47.0 (± 15.3), BST: 47.9 (± 15.2)	DO: 47.3 (± 12.7)	DO: 47.22 (± 15.61), BST: 48.23 (± 15.44)	ST: 32.45 (± 0.08)	45.3 (± 12)
Mean follow-up	Total thyroidectomy	3 years	14.5 years	10 years	20 months	5 years	NR	> 6 months
	Partial thyroidectomy	3 years	10 years	10 years	20 months	5 years	NR	> 6 months
Study duration		2003–2005	1975–1985	2000–2003	2009–2011	2001–2008	2014–1015	2007–2008
Treatment protocol	Total thyroidectomy	Extracapsular dissection with none remnant tissue	Tota unilateral/bilaterall extracapsular removal of the thyroid lobes and the pyramidal lobe	Extracapsular total thyroidectomy	Total thyroidectomy	Extracapsular total thyroidectomy	Total thyroidectomy	Extracapsular dissection
	Partial thyroidectomy	NTT: capsular dissection method, leaving less than 1 g of remnant tissue around the Berry ligament	ST: subtotal lobectomy was performed with contralateral subtotal resection leaving 3 ± 5 g (about 1.5 ± 2 cm3) on the less affected side	DO: less than 5 g homogeneous remnant was left unilaterally at either the superior pole or posterior depending on the location of detected nodules	DO: less than 5 g homogeneous remnant was left unilaterally at either the superior pole or posterior depending on the location of detected nodules	DO: unilateral extracapsular total thyroidectomy and contralateral subtotal thyroid lobe resection (leaving a thyroid stump of approximately 2 g of normal remnant tissue), BST: bilateral subtotal thyroidectomy (leaving on both sides of the neck thyroid stumps of approximately 2 g of normal remnant tissue each)	Subtotal thyroidectomy	NTT: Capsular dissection method, leaving less than 1 g of remnant tissue around the Berry ligament
Study population		Nontoxic euthyroid multinodular goiter	Unilateral and bilateral multinodular goiter	Bilateral non toxic multinodular goiter	Unilateral or bilateral multinodular goiter	Bilateral non toxic multinodular goiter	Multinodular goiter	Non-toxic multinodular goiter
Levothyroxine therapy		NR	1.5–2.25 ug/kg	TT: mean 100.59 µg/day BST: mean 77.16 µg/day	NR	TT: mean 100.59 µg/day BST: mean 77.16 µg/day	NR	NR
Criteria used for recurrences		NR	Physical examination or ultrasound scan showed nodular involvement or an enlargement of the residual thyroid remnant	Presence of hypoechoic or hyperechoic 60 nodular pattern at least 5 mm in diameter, identification of perinodular hypoechogenic or hyperechogenic ultrasound halo, and presence of an anechoic lesion with a reinforced posterior wall	If one of the following criteria: Volume of one lobe more than 9 ml for women and more than 12 ml for men or volume of both lobes more than 18 ml for women and more than 24 ml for men, Presence of a nodule more than 9 mm, Presence of an autonomous adenoma	Presence of hypoechoic or hyperechoic 60 nodular pattern at least 5 mm in diameter, identification of perinodular hypoechogenic or hyperechogenic ultrasound halo, and presence of an anechoic lesion with a reinforced posterior wall	NR	NR

Partial thyroidectomy is defined as near total thyroidectomy, bilateral subtotal shyroidectomy, or Dunhill operation (i.e., hemithyroidectomy plus subtotal sesection)

TT Total thyroidectomy, PT Partial thyroidectomy, DO Dunhill operation, BST Bilateral subtotal thyroidectomy, SD Standard deviation, NR Not reported

Risk of Bias Assessment and Quality of Evidence Assessment

Out of seven eligible RCTs, five had an overall low risk of bias, and two had some concerns. Figures 2 and 3 show the details of the risk of bias assessment of the included RCTs. The rating for all outcomes, according to GRADE criteria, was elaborated in Supplementary Table B.

Fig. 2.

Risk of bias graph

Fig. 3.

Risk of bias summary

Goiter Recurrence

Three articles representing two RCTs reported data on goiter recurrence (n = 711) [8–11]. The incidence of goiter recurrence was 0.76% in the TT group and 9% in the PT group (risk difference of 8.24%). TT showed significantly better results compared to PT in terms of goiter recurrence (RR = 0.05, 95% CI 0.02 to 0.13, P < 0.00001; I² = 0%, GRADE = moderate). Similarly, the subgroup analysis showed a significant difference between TT and STT (RR = 0.04, 95% CI 0.01 to 0.13, P < 0.00001; I² = 0%, GRADE = moderate), TT versus DO (RR = 0.08, 95% CI 0.02 to 0.36, P = 0.0008; I² = 0%, GRADE = moderate) (Fig. 4 and Supplementary Table B). The funnel plot was symmetrical on visual inspection, and no evidence of publication bias was noted (Supplementary Figure A).

Fig. 4.

Forest plot of goiter recurrence

Thyroid Cancer Incidence

Four RCTs reported data on thyroid cancer incidence (n = 910) [5, 8–10]. The incidence of thyroid cancer was 8.3% in the TT group and 7.9% in the PT group (risk difference of 0.4%). TT showed similar results compared to PT in terms of thyroid cancer incidence (RR = 1.09, 95% CI 0.76 to 1.57, P = 0.63; I² = 0%, GRADE = high). Similarly, the subgroup analysis showed no statistically significant difference between TT and STT (RR = 1.11, 95% CI 0.59 to 2.12, P = 0.74; I² = not applicable, GRADE = Moderate), TT versus DO (RR = 1.29, 95% CI 0.75 to 2.22, P = 0.36; I² = 0%, GRADE = Moderate), and TT versus NTT (RR = 0.77, 95% CI 0.36 to 1.66, P = 0.36; I² = not applicable, GRADE = Moderate) (Fig. 5 and Supplementary Table B). The funnel plot was symmetrical on visual inspection, and no evidence of publication bias was noted (Supplementary Figure B).

Fig. 5.

Forest plot of cancer incidence

Reported Adverse Events

Seven RCTs reported data on adverse events (n = 1483) [4–6, 8–11]. The most prevalent reported outcomes were temporary hypoparathyroidism and recurrent laryngeal nerve injury (18.98% and 5.27%, respectively). The other adverse events are provided in Supplementary Table C. The incidence of adverse events was 10.5%% in the TT group and 5.11%% in the PT group (risk difference of 5.39%). PT showed significantly better results compared to TT in terms of adverse events (RR = 2.18, 95% CI 1.64 to 2.89, P < 0.00001: I² = 63%, GRADE = high). Similarly, the subgroup analysis showed a significant difference that favors STT over TT (RR = 2.40, 95% CI 1.86 to 3.09, P < 0.00001; I² = 0%, GRADE = moderate), DO over TT (RR = 1.51, 95% CI 1.18 to 1.95, P = 0.001; I² = 0%, GRADE = moderate). On the contrary, there is no statistically significant difference between NTT and TT (RR = 7.83, 95% CI 0.89 to 68.94, P = 0.06: I² = 79%, GRADE = moderate) (Fig. 6, Supplementary Table B). The funnel plot was symmetrical on visual inspection, and no evidence of publication bias was noted (Supplementary Figure C).

Fig. 6.

Forest plot of adverse events

Discussion

Thyroid follicular nodular disease (MNG) is one of the most common surgical conditions during head and neck surgery. The surgical management of MNG remains controversial. There are several surgical procedures for MNG, including TT, STT, NTT, and DO [1, 2]. Marcin Barczynski et al. defined total thyroidectomy as extracapsular removal of the thyroid lobes and the pyramidal lobe, STT as unilateral or bilateral lobectomy with leaving approximately 2 g of normal remnant tissue, and DO as unilateral extracapsular total thyroid lobectomy and contralateral subtotal thyroid lobe resection with leaving a thyroid stump of approximately 2 g of normal remnant tissue [9]. According to Unalp et al., NTT is defined as capsular dissection of the thyroid lobes, leaving less than 1 g of remnant tissue around the Berry ligament [5]. Ozbas et al. claimed that NTT can be used as an alternative to TT since it offers a combination of TT advantages (i.e., no recurrences) with those of STT (i.e., low incidence of transient and permanent hypoparathyroidism) [18]. On the other hand, Pappalardo et al. suggested that there is no additional benefit that can be offered by NTT [10]. For this review, we divided the procedures into two groups, as this was done in previous studies; participants who underwent total thyroidectomy were assigned as the intervention group, and participants who underwent STT, NTT, or DO were classified as the comparison group (i.e., partial thyroidectomy) [12–14, 19]. In this study, we considered NTT as PT, as it was suggested by Si et al. [15]. The pooled effect estimate showed that TT was similar to PT in terms of cancer incidence and permanent complications, but it was superior to PT in terms of reducing goiter recurrence. In contrast, TT is inferior to PT in terms of transient complications. The subgroup analysis of all reported outcomes was done based on the operation type. Our review included three new RCTs involving a total of 939 participants, which increased the number of individuals available for the analysis more than in previous systematic reviews [7, 11, 12]. Goiter recurrence is considered a long-term complication of MNG surgical management. Some of the included studies considered the following criteria for goiter recurrence incidence: nodular involvement or enlargement of the residual thyroid remnant on either a physical examination or ultrasound scan [9, 10]. Limited procedures leave behind abnormal thyroid tissue and increase the potential risks of either persistent symptoms or recurrent disease in the future and require lifelong surveillance [20–23]. This was proven by the previous studies as well as our results showing that the risk of recurrence was significantly higher in the STT and DO groups than in the TT group [13, 14, 18]. No RCT was found to compare the recurrence rates of MNG after NTT to the TT group [5]. A higher recurrence rate carries an increased rate of surgical morbidity due to the need for re-operation [24–26]. The recurrence rate is largely dependent on the length of the follow-up period, i.e., the prevalence of recurrence increases as the time of follow-up increases, which is also supported by the outcomes of Marcin Barczynski et al. The rate of goiter recurrence after STT at a 5-year follow-up was 11.58%, whereas at a 10-year follow-up, it increased to 22.4%. Furthermore, Barczynski et al. showed that the rate of goiter recurrence after TT at 5-year follow-up was 0.52%, whereas at 10-year follow-up, it was 0.6%, and the recurrence rate after DO at 5-year and 10-year follow-up was 4.71% and 8.6%, respectively [9, 12]. Most cases of thyroid malignancy within MNG consisted of micro-papillary or micro-invasive follicular cancers [18, 27–29]. According to Barczynski et al., incidental thyroid cancer was diagnosed in 18 out of 191 patients (9.42%) in the TT group, compared to 15 out of 189 patients (7.94%) in the DO group and 16 out of 190 patients (8.42%) in the STT group [9]. In another study, out of 49 patients, encapsulated papillary microcarcinoma was encountered in 41 of them (83.66%), follicular cancer was reported in five patients (10.2%), and only three patients were diagnosed with Hurthle cell carcinoma (6.12%) [10]. Erbil et al. reported the incidence of papillary cancer as 9.6% in TT (10/104) and 12.5% in NTT (14/112) [6]. The incidence of thyroid cancer in all previous systematic reviews and meta-analyses was comparable between TT and PT [13, 14]. One of the previous systematic reviews and meta-analyses which compared TT versus NTT and STT did not report the cancer incidence, although it was mentioned in one of the included RCTs [6, 15]. In our systematic review and meta-analysis, TT showed similar results compared to PT in terms of thyroid cancer incidence (8.68% and 7.9%, respectively). All trials studying recurrent laryngeal nerve injury (RLNI) reported using a laryngoscopy examination to assess the function of vocal folds. Serum calcium levels were measured to assess the parathyroid function 24 h postoperatively. Mobayen et al. and Pappalardo et al. defined permanent RLNI as the vocal fold palsy that lasted for more than 6 months postoperatively [10, 11]. Hypocalcemia was defined as a fall in the total serum calcium level by less than 2.0 mmol/L, based on the research of Marcin Barczynski et al. as well as Pappalardo et al. [9, 10]. Permanent complications were defined as those lasting for 6 months or longer [9–11]. Surgeons continue to debate whether the potential benefits of a total thyroidectomy outweigh the potential complications. In the literature, the main reported postoperative complications are transient/persistent recurrent laryngeal nerve (RLN) palsy and transient/persistent hypoparathyroidism [4, 9, 30]. In our study, the most prevalent reported complication is temporary hypoparathyroidism (18.98%), followed by temporary RLNI (5.27%). The incidence of adverse events was 8.54% in the TT group and 3.74% in the PT group. PT showed significantly better results compared to TT in terms of adverse events. Similarly, the subgroup analysis showed a significant difference that favors STT and DO over TT. On the contrary, there is no statistically significant difference between NTT and TT; hence, NTT can be considered TT. The previous systematic reviews and meta-analyses showed that the incidence of transient complications of TT was significantly higher than that of NTT, STT, and DO. On the other hand, the results of the permanent adverse events in both groups were comparable. The reported incidence of transient RLNI and transient hypocalcemia after TT was 6.5% and 23.9%, respectively, whereas the incidence of transient RLNI and transient hypocalcemia after PT was 2.9% and 6.5%, respectively. The incidence of persistent RLNI and hypocalcemia after TT was 1.09% and 0.82%, respectively, and for persistent RLNI and hypocalcemia after PT, it was 0.8% and 0.26%, respectively [13–15]. However, it is important to keep in mind the possibility of RLN injury after PT in some patients, especially if the surgeons are not well-trained [21]. The risk of RLN paresis among various types of thyroid resection, besides its effects on quality of life for several weeks to months, warrants particular attention. The routine preoperative and postoperative laryngoscopic examination allows early detection of RLN injury. Also, intraoperative neural monitoring provides direct visualization of RLN during the operation, and it was found to be an effective technique that significantly lowered the incidence of RLN palsy [21, 30, 31]. In concordance with the previous studies, our systematic review and meta-analysis showed an insignificant difference between TT and PT in terms of permanent RLNI and hypoparathyroidism after TT (0.75% and 0.68%, respectively) and after PT (0.84% and 0.15%, respectively). On the contrary, PT has better results in terms of transient RLNI and hypoparathyroidism (4.48% and 15.4%, respectively) in comparison to TT (6.38% and 23.8%, respectively). This can be explained by the larger resection extent of TT compared to the limited resection in NTT, STT, and DO; hence, TT is associated with a higher risk of damaging the recurrent laryngeal nerves and parathyroid glands. Only one study reported hemorrhage after PT with an incidence of (0.76%) and after TT (0%) [9]. Furthermore, Mobayen et al. study is the only one that reported wound seroma as an adverse event following thyroid resection with an incidence of 3.3% after TT and 2.4% after PT [11]. Although our data agree with the literature, there is good evidence to demonstrate that with well-trained surgeons and the use of the correct surgical technique, total thyroidectomy can be performed with extremely low complications, particularly by using capsular dissection techniques, the identification and preservation of RLN, and the blood supply to the parathyroid glands [32–34]. Furthermore, Ozbas et al. concluded that qualified and experienced surgeons with good operative skills determine the morbidity rates rather than the type of operation [18]. Although the popularity of STT for MNG is decreasing with time, most studies agree that there is no difference between the various techniques of partial thyroidectomy in terms of postoperative hemorrhage, RLN injury, or hypoparathyroidism [4, 9, 33, 34]. The strength of this meta-analysis is the inclusion of the three new RCTs, which provided, up to our knowledge, the largest sample size of a systematic review comparing the desired outcomes of TT versus PT by adding 939 participants to the previous sample size [7, 11, 12]. This systematic review and meta-analysis is the first study to consider near-total thyroidectomy, subtotal thyroidectomy, and DO as partial thyroidectomy. As in the previous studies, they compared TT/NTT versus STT or DO whereas some other studies suggested that TT is different from NTT [13–15]. Our method is justified by the Akdeniz et al. study, which defined STT and DO as PT; furthermore, Si et al. considered NTT as PT [15, 19]. Our review was able to assess the quality of the provided evidence using GRADE criteria. The quality of the provided evidence was high for cancer incidence and adverse events, while it was moderate for goiter recurrence, and this is attributed mainly to the relatively small number of trials and participants. We acknowledge that our review has some limitations. The definition of a clear cut-off for the different procedures was inconsistent among the included studies. We decided to use 2 g of remnant thyroid tissue as the cut-off size for STT and DO and less than 1 g of remnant thyroid tissue as the cut-off size for NTT, which was defined by previous systematic reviews to differentiate between the types of thyroidectomy [5, 9]. Another limitation is the limited number of studies comparing TT to PT with an overall short follow-up duration, although Barczynski et al. had a relatively long-term follow-up period of 10 years that objectively estimates the effects of different thyroid resection surgeries [9, 12].

Conclusion

For treating MNG with thyroidectomy, TT is considered a feasible and safe procedure with a significantly lower risk of goiter recurrence. It is comparable to PT in terms of persistent adverse events but has a higher risk of transient adverse events. There is no statistically significant difference between TT and PT in terms of thyroid cancer incidence.

Author Contributions

D.A.K. contributed to the research idea, data extraction, statistical interpretation, risk of bias assessment, the certainty of evidence assessment, creating the tables, and writing the manuscript. J.F.A. contributed to the research idea, data extraction, and risk of bias assessment. M.A.K. contributed to the design/protocol of the study, creating the data extraction model, statistical analysis/interpretation, and certainty of evidence assessment. A.H.A. and A.A.A. contributed to the data extraction and risk of bias assessment. B.I.A. and M.A.G. contributed to the discussion of the manuscript. H.A.H. is the senior author who supervised the project and reviewed the manuscript. All the authors have read, edited, and approved the final manuscript.

Data Availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethical Approval

Not required.

Conflict of Interests

The authors declare no competing interests.