Prognostic Assessment of BRAF Mutation in Preoperative Thyroid Fine-Needle Aspiration Specimens

Rita Abi-Raad, MD; Manju L. Prasad, MD; Jingwei Zheng, MD; Pei Hui, MD, PhD; Berrin Ustun, MD; Kevin Schofield, CT(ASCP); Guoping Cai, MD; Adebowale J. Adeniran, MD


Am J Clin Pathol. 2021;156(1):100-108. 

In This Article


Clinical Features and BRAF Mutation

The patients included 122 (81%) women and 29 (19%) men with a median age of 49 years (range, 8–63 years) Table 1. The cases included 69 FNAs from the right lobe, 66 from the left lobe, 14 from the isthmus, and 2 from cervical LNs. BRAF mutation testing was performed on 32 cases of FLUS (category III), 30 cases of suspicious for PTC (category V), and 88 cases of positive for PTC (category VI). BRAF testing was performed on one case classified as follicular neoplasm per clinician request. Seventy (46%) cases were BRAF mutant, and 81 (54%) were wild-type. BRAF mutation was identified in 4 (12.5%) FLUS, 9 (30%) suspicious, and 57 (64.8%) positive for PTC cytology diagnosis cases, respectively (P < .001).

All patients subsequently underwent a surgical procedure for definitive diagnosis and treatment; 143 patients underwent total thyroidectomy, and 8 patients had a lobectomy. Total thyroidectomy was performed in 99% (69/70) of BRAF mutant tumors and 91% (74/81) of wild-type cases. Total thyroidectomy was more frequently performed in patients with a positive diagnosis of PTC on cytology, who were also more likely to harbor a BRAF mutation (65%; 57/88), compared to patients with a FLUS or suspicious diagnosis (21%; 13/62). Among patients with total thyroidectomy, 134 had initial thyroidectomy (51% BRAF positive and 48% BRAF negative), and 9 had initial lobectomy followed by completion thyroidectomy (all BRAF negative); 8 cases were follicular variant of papillary thyroid carcinoma (FVPTC) diagnosed as FLUS, and 1 case was a sclerosing variant of PTC, diagnosed as suspicious for PTC on FNA. As shown in Table 1, there were no differences in patients' age, sex, and LN dissection between BRAF mutant and wild-type groups.

BRAF Mutation and Histopathologic Features

The tumors ranged from 0.1 to 5.7 cm in size (mean, 1.5 cm) and showed no association with BRAF mutation (1.55 ± 1.23 cm in mutant vs 1.47 ± 1.18 cm in wild-type, P > .05) Table 2. The histologic subtypes of PTCs were as follows: 50 classic variant, 31 follicular variant, 9 tall-cell variant, 48 micro-PTC, and 13 other types. On reviewing preoperative ultrasound and final thyroidectomy specimen reports, we found that the nodule harboring micro-PTC was actually biopsied in 48 cases. Only those micro-PTC cases were included in the study (mean size, 0.58 ± 0.19 cm). BRAF mutation examined in FNA specimens, in comparison with wild-type BRAF, strongly predicted PTC histologic subtype. Classic and tall-cell variants of PTC were more likely to harbor the mutation (64% [32/50] and 89% [8/9], respectively) compared to FVPTC (19%; 6/31), while FVPTC and micro-PTC were more likely to be wild-type (80.6% [25/31] and 62.5% [30/48], respectively) (P < .001) (Table 2).

ETE was present in 15 cases: gross ETE as described at the time of surgery was present in 7 cases (6 mutant and 1 wild-type BRAF), and microscopic ETE was present in 8 cases (6 mutant and 2 wild-type BRAF). For the purpose of this study, any ETE (microscopic or gross) will be referred to as ETE. The presence of BRAF mutation was significantly associated with PTC subtype (P < .001), ETE (P = .006), and T stage (P = .04). Since only gross ETE is considered in the AJCC eighth edition staging manual,[27] a repeat analysis considering only gross ETE revealed a statistically significant association between BRAF mutation and gross ETE (P < .001). The tumors with BRAF mutation did not seem to increase the incidence of LVI (P > .05) (Table 2). Although mutant BRAF tumors were significantly associated with ETE, gross ETE, and higher tumor stage, BRAF mutation was also more likely to be encountered in histologic variants with high-risk pathologic features such as classic and tall-cell variant of PTC. ETE was present in 7 classic (6 mutant and 1 wild-type BRAF) and 6 tall-cell variants of PTC (5 mutant and 1 wild-type BRAF), while there was no ETE detected in the remaining 43 classic (26 mutant and 17 wild-type BRAF) or 3 tall-cell variants of PTC (all mutant BRAF). Gross ETE was present in 2 classic (both mutant BRAF) and 5 tall-cell variants (4 mutant and 1 wild-type BRAF). The remaining 48 classic (30 mutant and 18 wild-type) and 4 tall-cell variants of PTC (all mutant BRAF) showed no gross ETE. No ETE was identified in FVPTC. While a subsequent analysis within each histologic subtype revealed no significant association between BRAF mutation and the presence of ETE, gross ETE, and higher tumor stage, this analysis falls on numbers that are too small for a definitive conclusion.

BRAF Mutation and Lymph Node Metastasis

Preoperative LN cytologic assessment was performed in 35 patients using neck ultrasound-guided FNA. Suspicious/positive LN metastases were identified in 5 (29%) of 17 BRAF mutant and in 9 (50%) of 18 BRAF wild-type patients. LN dissection was performed at a slightly higher frequency in BRAF mutant (87%; 61/70) compared with wild-type BRAF (74%; 60/81), but the difference was not significant (P > .05). Central neck LN dissection was performed in 120 (79%) patients, more frequently in mutant BRAF (86%) compared with wild-type BRAF (74%) (P > .05). We did not observe the same tendency in lateral LN neck dissection, which was performed in 12 (8%) patients: 7% with mutant BRAF and 9% wild-type BRAF. Eleven patients had both central and lateral compartments dissected. The mean number of LNs removed was 8 in mutant BRAF and 12 in wild-type BRAF. The proportions of LN metastases were 12% and 13% in the central and lateral compartments, respectively. None of the patients had lateral LN metastasis in the absence of central LN metastasis. Of mutant BRAF, 33% (20/60) and 80% (4/5) of patients had positive LN in the central and lateral compartments, respectively, while 35% (21/60) and 86% (6/7) of wild-type BRAF patients had positive LN in the central and lateral compartments, respectively. There was no statistically significant difference in central (P = .847) or lateral (P = .793) LN metastasis between mutant or wild-type BRAF groups (Table 2). In patients younger than 55 years, the proportion of positive LNs was 37% (16/43) in BRAF mutant and 39% (18/46) in wild-type tumors. Similarly, in patients 55 years or older, there was no difference in LN metastasis between mutant (5/18; 28%) and wild-type BRAF (3/14; 21%). In both age groups, BRAF status was not associated with central or lateral LN metastasis Table 3.

Clinical Follow-up

In total, 102 patients had thyroid cancer-specific follow-up with a mean of 63 months (range, 4–107), 21 patients were lost to follow-up, and 28 patients had no thyroid cancer–specific follow-up. Of the 77 patients with available postoperative 131I treatment information, 70% (24/34) of patients with mutant BRAF received 131I treatment compared with 63% (27/43) of patients with wild-type tumors (P = .628). Of patients with available 131I treatment information, all patients (n = 27) with nodal involvement received 131I treatment compared with 53% (21/40) of patients without nodal involvement (P < .001). Only three (3%) patients had local recurrence. BRAF mutation did not predict recurrence of PTC. Specifically, one of the patients who recurred had mutant (no LN metastasis at thyroidectomy) and two had wild-type BRAF (one patient with and one patient without metastasis at thyroidectomy). All three patients had postoperative 131I treatment. Of 31 patients with no LN dissection performed, 15 had available follow-up and had no recurrence (mean follow-up, 67 months).