Clinical Significance of Gross Extrathyroidal Extension to Only the Strap Muscle According to Tumor Size in Differentiated Thyroid Cancer: A Systematic Review and Meta-Analysis
Article information
Abstract
Objectives.
The presence of extrathyroidal extension (ETE) in patients with differentiated thyroid cancer (DTC) serves as a significant prognostic indicator. Consequently, the staging of DTC is categorized into extensive ETE and gross ETE that solely impacts the strap muscles (gross strap muscle invasion [gSMI]). However, there is a lack of sufficient evidence concerning the relationship between gSMI and prognosis, particularly in terms of tumor size.
Methods.
Relevant literature was searched in Medline, Embase, Cochrane Library, and KoreaMed. All procedures were conducted in accordance with PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines and carried out by two independent reviewers. The meta-analysis utilized a random-effects model to account for the diversity of the studies. Risk of Bias for Nonrandomized Studies (RoBANS) version 2.0, an evaluation tool for non-randomized studies, was employed to assess the quality of the selected research. Clinical data from observational studies that examined the relationship between the degree of ETE and prognosis were gathered, and a meta-analysis was conducted.
Results.
Eighteen observational studies were included in this analysis. Subgroup analyses were conducted for each outcome. The findings revealed that the recurrence rate (odds ratio [OR], 2.498), disease-specific mortality (risk ratio [RR], 2.984), overall mortality (RR, 1.361), and lymph node (LN) metastasis (OR, 5.355) were significantly higher in patients with gSMI than in those without ETE. However, when the analysis was restricted to tumors measuring 4 cm or smaller, no significant differences in prognostic outcomes were observed, with the exception of LN metastasis.
Conclusion.
gSMI negatively impacts prognosis; however, this correlation diminishes with smaller tumor sizes. Thus, a more cautious approach is warranted during the treatment process.
INTRODUCTION
In thyroid cancer, extrathyroidal extension (ETE) refers to the growth and invasion of the tumor into surrounding tissues beyond the thyroid capsule. The 2015 management guidelines for thyroid nodules and differentiated thyroid cancer (DTC), published by the American Thyroid Association, identify ETE as a major prognostic factor, associated with a recurrence rate ranging from 10% to 40% [1]. Consequently, the extent of ETE is considered a critical factor in determining the cancer stage in DTC.
Until the 2018 publication of the revised staging for DTC by the American Joint Committee on Cancer (AJCC) in 2018 (eighth edition), the extent of ETE was categorized into minimal and extensive, which includes the extension of the primary tumor to the trachea, esophagus, recurrent laryngeal nerve, larynx, subcutaneous soft tissue, skin, internal jugular vein, or carotid artery (seventh edition). This classification was the basis for categorizing tumor stages as T3 (tumor size >4 cm or minimal ETE [sternothyroid muscle, perithyroid soft tissues]) and T4 (gross ETE from the tumor at any size) [2]. Minimal ETE is pathologically defined as the extrathyroidal involvement of perithyroid adipose tissue and strap muscles (sternohyoid, sternothyroid, omohyoid muscle) [2,3]. However, the concept of minimal ETE was found to have a weak correlation with prognosis compared to grossly confirmed ETE, as demonstrated by several studies [3,4]. Consequently, in the 2018 revision of the AJCC cancer staging manual (eighth edition), T3b in DTC was redefined to include only tumors with gross ETE limited to the strap muscles (gSMI), specifically excluding minimal ETE. Despite these changes, there remains a lack of comprehensive studies and definitive evidence on whether gSMI significantly impacts prognosis.
In this study, we conducted a systematic review and metaanalysis to evaluate the impact of gSMI on prognosis. Additionally, we sought to determine how tumor size influences the prognosis of patients with gSMI by categorizing tumor sizes and comparing prognostic outcomes within this group.
MATERIALS AND METHODS
Search strategy
The overall literature search and analysis process adhered to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines [5]. This study was registered with PROSPERO (registration No. CRD42022340174) at the beginning of the research period. Following the development of a search strategy, we conducted a comprehensive literature search using four databases: Medline, Embase, Cochrane Library, and KoreaMed. We included studies published from 1974 up to January 3, 2024. To enhance the search, we broadened the range of synonyms for terms related to the topic and devised a search strategy that incorporated all these terms. We also reviewed controlled vocabularies associated with the search terms in MeSH (Medical Subject Headings) and EMTREE (Embase Subject Headings), ensuring the inclusion of all pertinent terms. The detailed search strategy employed in this study can be found in Supplementary Table 1. There were no restrictions regarding language, study type, species, sex or age in the formulation of the search strategy. Additionally, gray literature and abstracts were considered.
Eligibility criteria
After applying the search strategy to four databases, a total of 819 documents were searched as of January 3, 2024. Including two additional papers obtained through hand-searching, the total number of papers retrieved was 821. Out of these, 517 papers were identified after removing duplicates (Fig. 1). The review process was carried out in two stages by two independent reviewers (HRW, BSK). In the first stage, 81 relevant documents were selected based on their titles and abstracts. In the second stage, a thorough review of the full content of these 81 papers was performed, resulting in 18 related papers being ultimately identified [6-23]. The reasons for excluding papers are listed in Fig. 1. The most common reason for exclusion was a mismatch between the main questions of this study and the content of the papers. Each step of the review was independently executed by the reviewers, with any disagreements resolved through discussion and consensus between them.
Flow diagram showing the selection of studies for inclusion in the systematic review and meta-analysis.
Data extraction
Data relevant to the objectives of this study were extracted by one reviewer (HRW), focusing on the prognosis of DTC. The participants were categorized into three groups: (1) the group without ETE (no-ETE group); (2) the group with ETE confirmed visually or radiologically only in the strap muscles (gSMI group); (3) and the group with invasion of surrounding organs, excluding the strap muscles (extensive ETE group). The extracted outcomes included the recurrence rate, recurrence-free survival (RFS), disease-specific survival (DSS), overall survival (OS), and lymph node (LN) metastasis rate.
Risk of bias analysis
All 18 papers included in this study were confirmed to be retrospective observational studies. Consequently, the quality of the studies incorporated into the meta-analysis was assessed using the Risk of Bias for Nonrandomized Studies (RoBANS) version 2.0, a tool designed for evaluating the quality of non-randomized experimental studies. Eight items were evaluated, with the risk levels categorized as high, low or unclear. The risk of bias evaluation was independently conducted by two reviewers (HRW, JWK). In cases where their assessments differed, the evaluation indicating the higher risk was chosen. The results of the risk of bias assessment are presented in Supplementary Fig. 1. These findings were summarized using Review Manager 5.4 (Cochrane). All meta-analysis results showed heterogeneity with I2 statistics from 0.000% to 95.749%.
Quantitative data analysis
Data on prognosis related to the extent of ETE were gathered for patients with DTC. All data were analyzed using the Comprehensive Meta-Analytic program version 4 (Biostat Inc.). The recurrence rate and LN metastasis rate were reported as odds ratios (ORs), while the results for RFS, DSS, and OS were reported as risk ratios (RRs). All meta-analyses were conducted using 95% CIs and employed a random-effects model.
RESULTS
Study selection
After applying the exclusion criteria, 18 papers were ultimately selected. The complete study selection process and the exclusion criteria are detailed in Fig. 1. Table 1 summarizes the studies included in the analysis, listing sample numbers for each group, relevant patient information, and the necessary outcomes and data [6-23].
Studies included in the systematic review and meta-analysis
Comparison of the recurrence rate between gSMI and other groups
First, we analyzed the recurrence rates between the group without ETE and the group with gSMI, including data from seven studies [8,12,18-20,22,23]. The gSMI group comprised 1,354 patients, with 184 (13.6%) experiencing recurrence. In contrast, among the 11,609 patients in the no-ETE group, 721 (6.2%) had recurrences. A meta-analysis using a random effects model showed that the OR of recurrence for patients with gSMI was 2.498 (95% CI, 1.418–4.402; P=0.002) compared to those without ETE (Fig. 2A). Conversely, the OR for recurrence in the gSMI group was 0.618 when compared to the extensive ETE group, indicating a significantly lower recurrence rate (95% CI, 0.458–0.834; P=0.002) (Fig. 2B). This result was based on three studies [10,18,22]. Further analysis revealed an OR of 2.668 for the extensive ETE group compared to the no-ETE group, suggesting a higher recurrence rate in the extensive ETE group (95% CI, 1.660–4.286; P=0.000) (Supplementary Fig. 2) [18,22].
Recurrence rate in patients with gross extrathyroidal extension (ETE) only affecting the strap muscle compared with (A) patients without ETE and (B) those with extensive ETE.
Comparison of RFS, DSS, and OS between gSMI and other groups
A meta-analysis was performed on studies reporting outcomes for RFS, DSS, and OS, which are important factors related to the prognosis of DTC. The results of each analysis are expressed as an RR, indicating the relative likelihood of the event occurring within the specified period. Six studies reported comparisons of RFS between the no-ETE group and the gSMI group [6,8,18,20, 22,23], and the RR for recurrence in the gSMI group was 2.497 (95% CI, 1.509–4.131; P=0.000) compared to the group without ETE (Fig. 3A).
(A) Recurrence-free survival, (B) disease-specific survival, and (C) overall survival in patients with gross extrathyroidal extension (ETE) only affecting the strap muscle compared to patients without ETE.
Results for DSS were reported in four studies [11,13,21,22]. The OR for the gSMI group compared to the no-ETE group was 2.984, indicating a higher disease-specific mortality rate in the gSMI group (95% CI, 1.458–6.105; P=0.003) than in the noETE group (Fig. 3B). Similarly, OS results were reported in four studies [13,15,18,23], with the OR for the gSMI group compared to the no-ETE group being 1.361 (Fig. 3C). This suggests a higher overall mortality rate in the gSMI group (95% CI, 1.129–1.642; P=0.001) than in the no-ETE group. The study explored the impact of extensive ETE on prognosis compared to cases without ETE. Results indicated that extensive ETE was linked to a significantly increased risk of recurrence, disease-specific mortality, and overall mortality, mirroring the risks associated with gSMI (Supplementary Fig. 3).
Comparison of the regional LN metastasis rate between gSMI and other groups
A meta-analysis was conducted to assess the rate of regional LN metastasis. Nine studies documented LN metastasis following ETE [6,8,11,13,15,18,19,22,23]. In the gSMI group, LN metastasis was observed in 9,144 (48.5%) of 18,851 patients, whereas in the no-ETE group, it was reported in 33,065 (16.0%) of 206,796 patients. Thus, the gSMI group exhibited a significantly higher likelihood of LN metastasis (OR, 5.355; 95% CI, 4.216–6.801; P=0.000) than the no-ETE group (Fig. 4A).
Comparison of lymph node metastasis in patients with gross extrathyroidal extension (ETE) only affecting the strap muscle and without ETE. (A) Total lymph node metastasis, (B) N1a (central lymph node metastasis), and (C) N1b (lateral lymph node metastasis).
When classifying the LN metastasis stage of DTC according to the eighth edition of the AJCC system, central LN metastasis (level VI) is defined as N1a, and lateral neck LN metastasis is defined as N1b. A meta-analysis of studies reporting each stage separately showed that the gSMI group had an OR of 2.608 (95% CI, 1.755–3.876; P=0.000) for N1a disease (Fig. 4B) [6,8,11-14, 19,20,23] and an OR of 7.248 (95% CI, 3.648–14.398; P=0.001) for N1b disease (Fig. 4C) [6,8,11,13,19,23], reflecting a significant risk elevation for both stages compared to the no-ETE group.
When extensive ETE was present, the OR for LN metastasis was 6,278 (95% CI, 4.483–8.792; P=0.000) compared to the no-ETE group, corresponding to a significant increase in risk (Supplementary Fig. 4A) [6,11,13,15,18,22]. Additionally, a meta-analysis of six studies reporting LN metastasis in the gSMI group and the extensive ETE group [6,11,13,15,18,22] found that the gSMI group showed a lower rate of LN metastasis than the extensive ETE group, with an OR of 0.659 (95% CI, 0.520–0.836; P=0.001) (Supplementary Fig. 4B).
When analyzing distant metastasis, the gSMI group had a higher likelihood of distant metastasis (OR, 6.927; 95% CI, 5.894–8.141; P=0.000) than the no-ETE group (Supplementary Fig. 5A) [15,18]. However, gSMI was associated with a lower risk of distant metastasis than was observed with extensive ETE (OR, 0.458; 95% CI, 0.396–0.530; P=0.000) (Supplementary Fig. 5B) [15,18].
Comparison of prognostic parameters between gSMI and other groups in tumors measuring 4 cm or less
Next, we limited the tumor size to 4 cm or less to analyze the correlation between gSMI and prognosis, comparing it with other groups. Three studies reported on the recurrence rate [7,20,23]. In the gSMI group, recurrence occurred in 24 (5.7%) out of 420 patients, versus 64 (2.9%) out of 2,202 patients in the no-ETE group. The OR was 1.530 (95% CI, 0.807–2.940; P=0.193), showing no statistical significance between the two groups with regard to this outcome (Fig. 5A).
(A) Recurrence rate, (B) recurrence-free survival, (C) disease-specific survival, and (D) lymph node metastasis rate in patients with differentiated thyroid cancer measuring 4 cm or less with gross extrathyroidal extension (ETE) only affecting the strap muscle compared to patients without ETE.
For RFS, a meta-analysis of the results from three studies showed that the RR for recurrence in the gSMI group was 1.360 (95% CI, 0.844–2.191; P=0.207) compared to the no-ETE group [7,20,23], showing no statistically significant difference (Fig. 5B). Likewise, the RR for disease-related mortality was 1.893 (95% CI, 0.448–8.002; P=0.385) [16,21], which also did not reflect a statistically significant difference (Fig. 5C). However, the outcome of regional LN metastasis was significantly more common in the gSMI group than in the no-ETE group, with an OR of 3.013 (95% CI, 1.477–6.147; P=0.002) (Fig. 5D) [7,20,23].
DISCUSSION
First, to accurately determine the impact of gSMI on prognosis, we conducted a meta-analysis to compare the effects of gSMI with those of no ETE. Prior to this study, Zhang et al. [24] reported a similar meta-analysis that included six studies. This earlier analysis showed that DTC patients with gSMI experienced significantly higher rates of recurrence (OR, 1.46; 95% CI, 1.18–1.80; P=0.0004) and LN metastasis (OR, 4.19; 95% CI, 2.53–6.96; P<0.001) than those without ETE. However, the number of studies, especially those examining LN metastasis, was considered inadequate, as only three studies were included in the meta-analysis. Since 2020, more studies have been published on the prognosis of gSMI, including those utilizing big data, which warranted an updated meta-analysis. This meta-analysis incorporated 18 papers. The results demonstrated a significant increase in the recurrence rate among DTC patients with gSMI compared to those without ETE (OR, 2.498; 95% CI, 1.418–4.402; P=0.002). Additionally, when accounting for the follow-up period, the recurrence rate remained elevated (RR, 2.497; 95% CI, 1.509–4.131; P=0.000). Furthermore, LN metastasis was more prevalent in the gSMI group than in the no-ETE group (OR, 5.355; 95% CI, 4.216–6.801; P=0.000), with similar increases observed in both central neck LN and lateral neck LN.
The findings of both the existing study and this study are comparable. gSMI was linked to a poorer prognosis compared to the absence of ETE, yet it was associated with a better prognosis than extensive ETE. However, since tumor size significantly affects prognosis, the current staging system, which classifies stages based solely on the extent of ETE without accounting for tumor size, seems somewhat imprecise. Consequently, there is ongoing debate about the impact of tumor size on prognosis in patients with gSMI. According to Park et al. [7], there was no significant difference in RFS (RR, 0.556; 95% CI, 0.218–1.422; P=0.221) between cases where the tumor size was 2 cm or less with gSMI and the no-ETE group. However, for tumors measuring 2–4 cm, the gSMI group exhibited a higher risk of recurrence than the no-ETE group (RR, 2.659; 95% CI, 1.159–6.099; P=0.021) [7]. Meanwhile, Yoon et al. [16] reported that when the tumor size was limited to 4 cm or less, a multivariate analysis indicated no reduction in DSS in the gSMI group (RR, 2.659; 95% CI, 1.159–6.099; P=0.021). Therefore, our study aimed to assess the impact of tumor size on prognosis in the gSMI group by analyzing the influence of gSMI on prognosis in tumors 4 cm or smaller. Interestingly, when the tumor size was restricted to 4 cm or less, no significant differences were observed between the gSMI group and the no-ETE group in terms of recurrence rate (OR, 1.530; 95% CI, 0.807–2.940; P=0.193), RFS (OR, 1.360; 95% CI, 0.844–2.191; P=0.207), and DFS (OR, 1.893; 95% CI, 0.448–8.002; P=0.385) (Fig. 5A-C). Only the rate of LN metastasis was linked to a significant increase in risk (OR, 3.013; 95% CI, 1.477–6.147; P=0.002) (Fig. 5D). These results indicate that prognosis is not uniform across all patients with gSMI, suggesting that treatment strategies for this group should take into account additional factors, including tumor size.
However, this study has certain limitations regarding the selection of research subjects. Surgical treatment options for patients include thyroid lobectomy or total thyroidectomy, selected based on specific case factors such as whether a central neck LN dissection was performed or adjuvant radioactive iodine treatment was administered. Although the study incorporated 18 studies, the number of studies included in the meta-analysis varied in the subgroup analyses, ranging from a maximum of 7 to a minimum of 2, which limits the robustness of the evidence. Additionally, despite using a random-effect model in the meta-analysis to account for diversity between studies, heterogeneity varied from 0% to 95.7%, which weakened the power of the study’s results. However, the overall number of patients included in the analysis is relatively large compared to previous studies, and efforts were made to minimize selection bias by carefully selecting studies with potential overlaps in patient samples. Notably, this study is the first to analyze LN metastasis by stage (N1a and N1b) and to conduct a subgroup analysis based on tumor size specifically for patients with gSMI.
In conclusion, the presence of gSMI in DTC patients is associated with significantly higher risks of recurrence and mortality compared to those without ETE. This suggests that gSMI is a crucial factor in predicting patient prognosis. However, the impact of gSMI on prognosis may vary depending on the tumor size. Consequently, when gSMI is present but the tumor is small, treatment methods should be carefully chosen to enhance the patient’s prognosis, taking into account various clinical factors.
HIGHLIGHTS
▪ In differentiated thyroid cancer patients with only gross strap muscle invasion (gSMI), a clinical correlation with tumor size was not identified.
▪ gSMI was a significantly poor prognostic factor compared to the absence of extrathyroidal extension.
▪ In tumors ≤4 cm, there were no significant differences in prognosis except for lymph node metastasis.
▪ gSMI has a negative prognostic impact that depends on tumor size.
Notes
Bon Seok Koo is an editorial board member of the journal but was not involved in the peer reviewer selection, evaluation, or decision process of this article. No other potential conflict of interest relevant to this article was reported.
AUTHOR CONTRIBUTIONS
Conceptualization: HRW, BSK. Data curation: HOS, SY, JWC. Formal analysis: HRW, JWK. Funding acquisition: HRW, BSK. Methodology: HRW. Visualization: HRW. Writing – original draft: HRW. Writing – review & editing: HRW, BSK. All authors read and agreed to the published version of the manuscript.
Acknowledgements
This wowrk was supported by Chungnam National University Hospital (to HRW), the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (grant number: 2022R1C1C1008265 to HRW), a National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (grant No. RS-2024-00339029 to BSK) and the Patient-Centered Clinical Research Coordinating Center funded by the Ministry of Health & Welfare, Republic of Korea (grant No. HC19C0103).
SUPPLEMENTARY MATERIALS
Supplementary materials can be found online at https://doi.org/10.21053/ceo.2024.00162.
Search strategy
Risk of bias summary using Risk of Bias for Nonrandomized Studies (ROBANS) version 2.0 Tool: review authors’ judgements about each risk of bias item for each included study (red circle: high risk, green circle: low risk, yellow circle: unclear risk).
Recurrence rate in patients with extensive extrathyroidal extension (ETE) compared to patients without ETE.
(A) Recurrence-free survival (RFS), (B) disease-specific survival (DSS), (C) and overall survival (OS) in patients with extensive extrathyroidal extension (ETE) compared to patients without ETE.
Comparison of lymph node metastasis in patients with extensive extrathyroidal extension (ETE) versus those (A) without ETE or (B) with gross ETE only affecting the strap muscle.
Comparison of distant metastasis in patients with gross extrathyroidal extension (ETE) only affecting strap muscle versus (A) those without ETE and (B) those with extensive ETE.