Surgery · Medical Literature
Optimizing frozen section sampling for thyroid follicular neoplasms through intraoperative ultrasound guidance.
Abstract
To investigate the clinical value of intraoperative high-frequency ultrasound in assessing the capsular continuity of thyroid follicular neoplasms and to evaluate its feasibility in guiding optimal frozen section sampling during surgical resection to improve diagnostic accuracy. Patients who were scheduled for surgical resection for thyroid follicular neoplasms were prospectively enrolled and randomized into a control group or an experimental group. In the experimental group, the excised specimens underwent comprehensive intraoperative high-frequency ultrasonography to identify areas suspected of capsular invasion, thereby guiding precise sampling for intraoperative frozen section analysis. The value of this method was assessed using permanent histopathology as the gold standard. A total of 248 patients were included (130 in the control group and 118 in the experimental group). The diagnostic accuracy of intraoperative frozen section pathology was significantly greater in the experimental group than in the control group (91.53 vs 77.69%, P < 0.05). Intraoperative high-frequency ultrasonography can effectively identify abnormalities in tumor capsular continuity, guide optimized sampling for intraoperative frozen section analysis, significantly improve the intraoperative detection of malignant tumors, and provide reliable intraoperative guidance.
Clinical relevance
Does intraoperative high-frequency ultrasonography of excised specimens, used to identify areas of suspected capsular invasion and guide targeted frozen section sampling, improve the diagnostic accuracy of intraoperative frozen section pathology compared to conventional (non-guided) sampling in patients with thyroid follicular neoplasms? __SNAPSHOT__ {"studyDesign":"Prospective randomized controlled trial","setting":"Not reported (institution type and country not specified in abstract; authors affiliated with Chinese institutions based on author names and journal context)","sampleSize":"n=248 (control: 130, experimental: 118)","population":"Patients scheduled for surgical resection for thyroid follicular neoplasms; specific inclusion/exclusion criteria not reported in abstract","intervention":"Comprehensive intraoperative high-frequency ultrasonography of excised specimens to identify areas of suspected capsular invasion, with subsequent targeted frozen section sampling of those areas","comparator":"Conventional intraoperative frozen section sampling without ultrasound guidance (control group)","primaryOutcome":"Diagnostic accuracy of intraoperative frozen section pathology, validated against permanent (paraffin-embedded) histopathology as the gold standard","followUp":"Not reported (intraoperative/perioperative assessment only based on abstract)","fundingSource":"Not reported"}
Key findings
The primary outcome — diagnostic accuracy of intraoperative frozen section pathology — was significantly higher in the ultrasound-guided experimental group than in the control group (91.53% vs. 77.69%; P < 0.05). The absolute difference in diagnostic accuracy was approximately 13.84 percentage points (calculated). No relative risk, odds ratio, confidence intervals, NNT, or NNH were reported in the abstract. Sensitivity, specificity, positive predictive value, and negative predictive value were not individually reported. Specific rates of false-negative (missed malignancy) or false-positive results by group were not provided. Secondary outcomes, subgroup analyses (e.g., follicular adenoma vs. follicular thyroid carcinoma), and adverse events related to the intervention were not reported in the abstract.
Practice implications
This trial provides preliminary evidence that intraoperative high-frequency ultrasonographic examination of excised thyroid specimens can identify capsular abnormalities and guide targeted frozen section sampling, meaningfully improving intraoperative diagnostic accuracy over non-guided sampling (approximately 14 percentage points absolute improvement, calculated). The clinical importance of this finding is notable in the context of the well-documented limitations of conventional frozen section for follicular neoplasms — a diagnostic challenge where capsular and vascular invasion must be confirmed histologically and where routine frozen section has historically shown low sensitivity (as low as 3.4–10.5% for follicular thyroid carcinoma in prior literature). If confirmed, this approach could reduce the rate of deferred intraoperative diagnoses and potentially decrease the need for completion thyroidectomy as a second surgery. However, several important limitations prevent immediate practice change: only an abstract is available for full review, confidence intervals and individual diagnostic performance metrics (sensitivity, specificity, PPV, NPV) are not reported, and it is unclear how many of the 248 patients had malignant versus benign final diagnoses. The generalizability to centers without dedicated intraoperative ultrasound expertise or equipment is uncertain. These findings apply only to patients undergoing resection for follicular neoplasms and cannot be extrapolated to other thyroid pathologies. __LIMITATIONS__ 1. Only abstract available — full methodology, patient characteristics, randomization/allocation concealment details, blinding procedures, and complete statistical results cannot be assessed. 2. No confidence intervals reported for the primary outcome; statistical robustness cannot be fully evaluated. 3. Key diagnostic performance metrics (sensitivity, specificity, PPV, NPV, deferred result rate) are not individually reported, limiting clinical interpretation. 4. Sample size is modest (n=248) with an unexplained imbalance between arms (130 vs. 118), and no sample size justification or power calculation is described. 5. Institutional setting, operator experience with intraoperative ultrasound, and country are not specified, raising questions about generalizability and reproducibility. 6. Blinding of pathologists to the sampling method is not described, introducing potential assessment bias. 7. Follow-up duration is limited to the intraoperative/pathologic episode; long-term oncologic outcomes (recurrence, reoperation rates) are not reported. 8. Funding source is not reported, precluding assessment of potential conflicts of interest. 9. Proportion of malignant vs. benign final diagnoses within each group is not reported, making it impossible to assess whether the improved accuracy reflects better detection of malignancy or avoidance of false positives. 10. It is unclear whether the intervention adds meaningful operative time or resource burden. __BOTTOM_LINE_CATEGORY__ promising_but_preliminary __BOTTOM_LINE__ Intraoperative high-frequency ultrasonography of excised specimens to guide targeted frozen section sampling significantly improved diagnostic accuracy compared to conventional sampling in patients with thyroid follicular neoplasms (91.53% vs. 77.69%; P < 0.05; ~14 percentage points absolute improvement, calculated). These findings are clinically plausible and address a recognized diagnostic gap, but the absence of complete diagnostic metrics, confidence intervals, and full methodological transparency limits firm conclusions. Confirmation in larger, multi-center trials with full reporting of sensitivity, specificity, and long-term surgical outcomes is needed before this approach can be broadly adopted.