Oncology · Medical Literature
Risk factors associated with aggressive tumor phenotypes in papillary thyroid microcarcinoma: a systematic review and meta-analysis.
Abstract
In line with current guidelines, risk stratification in papillary thyroid microcarcinoma (PTMC) has shifted from simple diameter-based assessments to evaluation of individual tumor biology. Therefore, this study aimed to investigate how baseline clinical and molecular risk factors influence aggressive PTMC phenotypes. Registered in PROSPERO (CRD42024603662), we searched PubMed, Embase, and Cochrane through February 14, 2025. We analyzed risk factors for lymph node metastasis (LNM) and extrathyroidal extension (ETE) utilizing data from surgical cohorts of patients with PTMC, calculating pooled odds ratios (OR) and 95% confidence intervals (CI). A total of 274 articles we found increased odds of central lymph node metastasis (CLNM) associated with male sex (OR = 1.86; CI:1.73-2.00), age ( 5 and >7 mm OR = 2.26; CI:2.00-2.56 and OR = 1.81; CI:1.11-2.96), multifocality (OR = 1.96; CI:1.78-2.16), bilaterality (OR = 1.70; CI:1.45-2.00) and BRAFV600E mutation (OR = 1.51; CI:1.25-1.82). Lateral lymph node metastasis (LLNM) was associated with male sex (OR = 1.69; CI:1.51-1.90), age ( 5 mm and >7 mm, OR = 2.34; CI:1.68-3.25 and OR = 3.71; CI:3.15-4.36), multifocality (OR = 1.98; CI:1.46-2.70) and bilaterality (OR = 1.91; CI:1.22-3.00). For ETE, risk factors were as follows: tumor size (>5 mm OR = 3.00; CI:2.03-4.45), multifocality (OR = 2.41; CI:1.99-2.91), bilaterality (OR = 2.58; CI:1.26-5.27), and BRAFV600E mutation (OR = 2.19; CI:1.61-2.97). Male sex, age ( 5 mm), multifocality, bilaterality, and BRAFV600E mutation are clinically relevant risk factors for aggressive PTMC phenotypes. https://www.crd.york.ac.uk/PROSPERO/view/CRD42024603662, identifier CRD42024603662.
Clinical relevance
In patients with papillary thyroid microcarcinoma (PTMC, ≤10 mm), which baseline clinical and molecular risk factors are independently associated with aggressive tumor phenotypes—defined as central lymph node metastasis (CLNM), lateral lymph node metastasis (LLNM), or extrathyroidal extension (ETE)? __SNAPSHOT__ {"studyDesign":"Systematic review and meta-analysis of observational (predominantly retrospective) surgical cohort studies; PROSPERO-registered (CRD42024603662)","setting":"International; literature search of PubMed, Embase, and Cochrane databases through February 14, 2025; 80.1% of included articles originated from Asian populations","sampleSize":"274 articles included; >220,000 patients analyzed across 264 retrospective studies","population":"Patients with histopathologically confirmed PTMC (tumor ≤10 mm) who underwent surgical resection; predominantly single-center Asian surgical cohorts; pediatric cases may have been included in some contributing studies","intervention":"Evaluation of baseline clinical risk factors: male sex, younger age ( 5 mm and >7 mm thresholds), multifocality, bilaterality, and BRAFV600E mutation status","comparator":"Patients with PTMC without each respective risk factor (internal referent groups within each contributing study)","primaryOutcome":"Pooled odds ratios (OR) with 95% confidence intervals (CI) for three aggressive PTMC phenotypes: CLNM, LLNM, and ETE","followUp":"Not reported (cross-sectional surgical pathology outcomes; active surveillance outcomes not assessed)","fundingSource":"Not reported"}
Key findings
A total of 274 articles encompassing >220,000 PTMC surgical patients were included. For central lymph node metastasis (CLNM), statistically significant increased odds were observed for: male sex (OR=1.86; 95% CI: 1.73–2.00), younger age at cutoffs of 5 mm (OR=2.26; 95% CI: 2.00–2.56) and >7 mm (OR=1.81; 95% CI: 1.11–2.96), multifocality (OR=1.96; 95% CI: 1.78–2.16), bilaterality (OR=1.70; 95% CI: 1.45–2.00), and BRAFV600E mutation (OR=1.51; 95% CI: 1.25–1.82). For lateral lymph node metastasis (LLNM), significant associations were identified for: male sex (OR=1.69; 95% CI: 1.51–1.90), younger age at 5 mm (OR=2.34; 95% CI: 1.68–3.25) and >7 mm (OR=3.71; 95% CI: 3.15–4.36), multifocality (OR=1.98; 95% CI: 1.46–2.70), and bilaterality (OR=1.91; 95% CI: 1.22–3.00). Notably, BRAFV600E mutation was not reported as a significant predictor of LLNM in the abstract. Specific ETE results and p-values were not reported in the provided abstract. Heterogeneity was present across analyses (e.g., I² for CLNM by sex: 49%, 95% CI: 36–59%; for LLNM by sex: 57%, 95% CI: 40–69%); pooled adjusted ORs for LLNM by male sex lost statistical significance (OR=1.39; 95% CI: 0.96–2.01), indicating potential confounding. NNT/NNH are not calculable from aggregate OR data without baseline risk denominators. No adverse events are applicable to a risk-factor meta-analysis. Individual p-values for each pooled estimate were not reported in the abstract.
Practice implications
This large PROSPERO-registered meta-analysis establishes that multiple readily assessable preoperative clinical characteristics—male sex, younger age, larger intramicrocarcinoma tumor size (>5 mm and especially >7 mm), multifocality, and bilateral disease—are consistently and significantly associated with both CLNM and LLNM in surgically treated PTMC. BRAFV600E mutation adds independently to CLNM risk but its role for LLNM was less certain. Notably, tumor size >7 mm carried the strongest association with LLNM (OR=3.71), suggesting a gradient of nodal risk well within the microcarcinoma size range that challenges the adequacy of a simple 10 mm threshold for risk stratification. These findings are directly relevant to patient selection for active surveillance versus surgery, and for surgical extent decisions (e.g., prophylactic central neck dissection). However, several important caveats limit direct translation: (1) All contributing data derive from surgical cohorts, creating selection bias—patients managed with active surveillance are entirely unrepresented, and the absolute risks of LNM in unselected PTMC cannot be inferred. (2) The predominance of Asian cohorts (80.1%) limits generalizability to Western, iodine-replete, or ethnically diverse populations. (3) Risk factors were largely evaluated in univariate analyses; pooled adjusted ORs, where available, sometimes attenuated or lost significance (e.g., adjusted OR for LLNM by male sex: 1.39; 95% CI: 0.96–2.01, not statistically significant), reflecting confounding among co-occurring risk factors. (4) The study establishes associations, not independent predictive models suitable for direct clinical scoring. (5) The abstract does not report ETE-specific quantitative results, histologic subtype analyses, or disease-specific survival data. Only the abstract was available for this summary. __LIMITATIONS__ Overwhelming predominance of retrospective single-center studies from Asian populations (80.1%), substantially limiting generalizability to non-Asian patients and healthcare systems with different surgical thresholds and iodine status.,All data derived from surgical cohorts; patients managed with active surveillance are entirely excluded, introducing substantial selection bias and precluding conclusions about natural history of PTMC in non-operated patients.,Significant between-study heterogeneity (e.g., I²=49–57% for sex-CLNM/LLNM associations), with multiple potential sources including variability in lymph node dissection extent, outcome definitions (e.g., microscopic vs. gross ETE), iodine intake, and molecular testing methodology.,Pooled adjusted ORs, where reported, attenuate compared to unadjusted ORs and in some cases cross the null (e.g., adjusted OR for LLNM by male sex: 1.39; 95% CI: 0.96–2.01), raising concerns about residual confounding and the validity of individual factor estimates when co-occurring risk factors are present.,Histologic subtypes not analyzed separately due to inconsistent reporting, despite known differences in aggressive behavior across PTMC variants.,Pediatric cases may have been included in some contributing studies, potentially inflating risk estimates for younger-age categories.,No survival, recurrence, or disease-specific outcome data reported; aggressive phenotype surrogates (LNM, ETE) are pathologic endpoints, and their long-term clinical impact—particularly microscopic CLNM—remains debated.,Funding source not reported; potential for publication bias in the constituent literature, which typically over-represents high-risk or surgically complex PTMC cohorts.,Abstract-only availability for this summary; full data tables, subgroup analyses, and ETE-specific quantitative results could not be independently verified. __BOTTOM_LINE_CATEGORY__ potentially_practice_informing __BOTTOM_LINE__ This large meta-analysis of >220,000 surgically treated PTMC patients confirms that male sex, younger age, tumor size >5–7 mm, multifocality, bilaterality, and BRAFV600E mutation are each associated with increased odds of nodal metastasis, with tumor size >7 mm carrying the strongest LLNM signal (OR=3.71). These risk factors, aligned with evolving ATA guidelines, support multi-parametric rather than size-alone preoperative risk stratification. However, because all data originate from surgical cohorts with marked Asian predominance, these associations should inform—but not independently dictate—management decisions pending prospective validation in diverse, non-operatively managed populations.