BRAF p.V600E mutation is the most common molecular driver identified in papillary thyroid carcinomas (PTCs, 58.5–86.8%) [1, 6]. The mutated protein can be detected in the tumor cytoplasm using a mutation-specific antibody clone VE1, with excellent specificity (99–100%) and sensitivity (84–100%) across various tumor types, including thyroid cancers [7,8,9,10]. In our experience, BRAF V600E IHC is highly useful in highlighting the tumor extent and the unique patterns of invasion (lateral tubular growth and isolated tumor clusters), which are characteristic of BRAF-mutated PTCs (Fig. 1A and B) [11,12,13]. Identification of these patterns of invasion may have potential clinical significance as prior investigators have found them to be predictive of lymph node metastasis [11]. Furthermore, in well-encapsulated or well-defined thyroid tumors with PTC-like nuclear changes, BRAF V600E IHC can serve as an efficient diagnostic tool, as the presence of a BRAF p.V600E mutation is a desirable exclusionary criterion for the diagnosis of non-invasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTP) [14].

BRAF V600E IHC in diagnosing PTCs. A BRAF IHC highlights the tumor extent of this classic PTC (20 ×). B It also highlights lateral tubular growth and isolated tumor clusters, which are unique patterns of invasion most frequently encountered in BRAF p.V600E-mutated PTCs (100 ×). C Collision of a BRAF p.V600E-mutated PTC (left of the dotted line) and a RET::CCDC6 fusion-associated PTC (H&E, 50 ×). D BRAF IHC highlights the interface of the molecularly discordant tumors with variable intermingling of the neoplastic follicles in the center (200 ×). E Admixed lymph node metastasis consisting of molecularly discordant neoplastic microfollicles of BRAF p.V600E and RET::CCDC6 fusion-associated PTCs (H&E, 100 ×). F BRAF IHC highlights the BRAF-positive neoplastic follicles, confirming the admixture of the molecularly discordant tumors in metastasis (100 ×)

Previous research has attempted to distinguish intraglandular tumor spreading of PTC from multicentric tumors by applying histologic criteria based on tumor size and morphology [12]. In our recent study, BRAF V600E IHC was utilized to identify multifocal PTCs with discordant molecular drivers (BRAF-positive and BRAF-negative tumors), supporting diagnoses of multicentric disease in these cases [13]. However, in multifocal PTCs where all foci harbor mutual BRAF p.V600E mutation, additional clonality studies are warranted to clarify whether these represent truly multicentric tumors or intraglandular tumor spreading.

The occurrence of subclonal BRAF expression and intratumoral heterogeneity in PTCs remains controversial. While some studies using molecular assays suggest that these phenomena are not uncommon [15,16,17,18,19], others, including TCGA, report them as exceedingly rare, if not nonexistent [1, 20, 21]. Additionally, rare cases of PTCs with concurrent multiple molecular drivers were thought to represent subclonal BRAF expression and intratumoral heterogeneity [1, 15, 16, 18, 21]. However, these studies were limited by methodologies that could not clarify whether these mutations occurred within the same tumor cells or in distinct tumor populations. In our previous studies, BRAF V600E IHC consistently demonstrated homogeneous cytoplasmic staining in BRAF p.V600E-mutant PTCs, except in rare cases of molecularly discordant PTCs in collision [6, 13]. In the latter cases, the advantage of next-generation IHC over traditional molecular assays is highlighted, as BRAF V600E IHC enabled the precise identification and demarcation of molecularly discordant PTCs (Fig. 1C and D) and distinguished their corresponding lymph node metastasis (Fig. 1E and F) [13]. This approach offers a novel insight into rare PTCs reported to have concurrent multiple molecular drivers, which are theoretically mutually exclusive. Likewise, it suggests that a subset of cases previously considered to represent BRAF subclonality or intratumoral heterogeneity may now be attributed to the possible collision of molecularly discordant PTCs [1, 13, 15, 16, 18, 21]. Ultimately, further studies are needed to better define and validate the methodologies for assessing BRAF subclonality. Nevertheless, rare cases of heterogeneous BRAF IHC expression may complicate histologic interpretation and influence treatment decisions. In such cases, meticulous correlation of staining patterns with morphologic features, adequate tumor sampling, review of the IHC protocol, and molecular confirmation are recommended. The accurate diagnosis of multifocal PTCs with discordant molecular drivers, including cases of molecularly distinct tumor collisions, is clinically relevant for selecting appropriate targeted kinase inhibitors, especially in advanced or radioiodine refractory disease, which occurs in 10–15% of PTCs [2, 22].

In the lymph nodes, BRAF V600E IHC can highlight metastatic isolated tumor cells or clusters, which could be missed on routine H&E evaluation (Fig. 2A and B). This has implications for accurate pathologic staging. In addition, the distinction between benign intranodal thyroid inclusions and metastatic thyroid carcinoma is controversial and poses a diagnostic challenge among pathologists. Traditionally, microscopic size, benign cytologic features, and absence of papillae, stromal desmoplasia, and psammoma bodies were regarded as morphologic features in favor of benign thyroid inclusions [23, 24]. However, in the absence of clinical and imaging findings of the thyroid, these morphologic features remain subjective and do not provide conclusive evidence for the non-neoplastic nature of the benign intrathyroidal follicles. BRAF V600E IHC serves as an objective tool to distinguish between benign intranodal inclusions and lymph node involvement of a BRAF-positive PTC. In our recent study applying BRAF V600E IHC, we found benign thyroid inclusions and metastatic deposits of PTC may co-exist in the same lymph node as separate aggregates (Fig. 2C and D) or as intimately admixed deposits (Fig. 2E and F), suggesting a potential biological link between these two [25]. This led us to further refine the diagnostic criteria for benign intranodal thyroid inclusions with the recommendation of BRAF V600E IHC for optimal sensitivity and specificity.

BRAF V600E IHC in assessing tumor deposits in lymph nodes. A Rare isolated tumor cells resembling histiocytes are present in the lymph node (H&E, 400 ×). B BRAF IHC highlights the isolated tumor cells (400 ×). C Thyroid follicles present in the lymph node of a patient with a primary BRAF p.V600E-mutant PTC (H&E, 100 ×). D BRAF IHC distinguishes the tumor deposits from nearby benign intranodal inclusions (100 ×). E Thyroid follicles with varying degrees of atypia are present in the lymph node of a patient with a primary BRAF p.V600E-mutant PTC (H&E, 200 ×). F BRAF IHC confirms that the deposit is an admixture of both metastatic PTC and benign intranodal thyroid inclusion (200 ×)

Lastly, it should be emphasized that the utility of BRAF VE1 IHC is limited to detecting only BRAF p.V600E-mutant PTCs. It does not identify PTCs with non-p.V600E BRAF mutations, including other class I mutations (p.V600K/M/R/D), class II and III mutations, or BRAF fusions [26].

Mutations in the RAS family of oncogenes (NRAS, HRAS, and KRAS) are key molecular events in follicular-patterned neoplasms including follicular adenoma (FA, 20–25%), follicular thyroid carcinoma (FTC, 30–45%), and follicular variant of PTC (FVPTC, 30–45%), as well as high-grade tumors such as poorly differentiated thyroid carcinoma (20–40%) and anaplastic thyroid carcinoma (ATC, 20–30%) [27]. Similarly, NIFTP is found to be enriched with similar RAS mutations [28]. Furthermore, RAS mutations have also been found in thyroid nodules that would otherwise have been classified as hyperplastic based on morphology. From a diagnostic standpoint, these nodules could be more appropriately classified as FAs, given the presence of a driver mutation. However, in the absence of molecular information, distinguishing FA from these “hyperplastic-appearing” nodules remains largely subjective, especially in the setting of multinodular goiter. Importantly, this distinction has no clinical relevance following complete surgical resection [29]. Reflecting this evolving understanding, the latest WHO Classification of Endocrine and Neuroendocrine Tumors now favors the term thyroid follicular nodular disease over nodular hyperplasia or adenomatous hyperplasia [14]. This shift acknowledges that clonal nodules with oncogenic mutations can arise within multinodular goiter and may represent true neoplasms. Among the different RAS genes, mutations in NRAS are the most predominant (65–66%), followed by mutations in HRAS (21–27%) and KRAS (8–13%) [1, 30, 31]. The most frequent mutation is the substitution of glutamine by arginine in codon 61 (p.Q61R) of the RAS genes, representing 79–85%, 79–88%, and 25–60% of the NRAS, HRAS, and KRAS mutations, respectively [1, 31].

The pathologic classification of follicular-patterned tumors remains problematic as significant interobserver variability exists regarding the morphologic distinction of NIFTP and invasive encapsulated FVPTC from FA and FTC, and whether this pursuit still has clinical relevance [32]. More importantly, the dichotomization of PTCs by the TCGA project into RAS-like and BRAF-like PTCs confirms that most invasive encapsulated FVPTC and NIFTP are within the same family of RAS-like tumors and should be distinguished from the BRAF-like PTCs because of different tumor morphology, biological behavior, and response to radioiodine therapy [1]. This implies that in morphologically ambiguous follicular-patterned neoplasms, identifying the RAS mutation using next-generation IHC, in lieu of molecular analysis, offers a powerful tool in classifying these tumors into more clinically relevant groups and help obviate the existing subjectivity inherent in its morphological interpretation. Furthermore, once a RAS-like follicular-patterned neoplasm is identified, attention to its architecture becomes paramount to assess capsular and/or angioinvasion, which the IHC can help highlight, further improving diagnostic accuracy [1, 29].

The commercially available antibody SP174 effectively detects the RAS p.Q61R protein, demonstrating homogeneous moderate to strong granular cytoplasmic and/or membranous staining in tumor cells (Fig. 3A and B). It demonstrates high sensitivity (90.6%) and specificity (92.3%) for identifying Q61R mutations across all three RAS genes [29, 33,34,35]. In cases where non-specific cytoplasmic staining occurs in the background thyroid parenchyma, the presence of a membranous staining pattern can serve as a potential distinguishing feature to confirm that the tumor harbors clonal RAS p.Q61R mutation. In diagnosing follicular-patterned carcinomas, the IHC is valuable in delineating tumor border, infiltration, capsular and/or vascular invasion, and multifocality [33, 34]. In our experience, RAS Q61R IHC has been indispensable for confirming the diagnosis of minimally invasive FTC (Fig. 3C and D). Occasionally, follicular adenomas and hyperplastic nodules within follicular nodular disease can grow in close proximity, making it extremely challenging to distinguish them morphologically from an FTC with an invasive satellite nodule. In such cases, RAS Q61R IHC can be a valuable diagnostic tool (Fig. 3E and F). Similarly, the IHC can also highlight the multinodular infiltrative growth pattern characteristic of widely invasive FTCs with mixed macro-microfollicular architectural patterns, which can be mistaken for adenomatoid nodules within follicular nodular disease (Fig. 4A and B). Furthermore, RAS Q61R IHC can confirm the neoplastic nature of metastatic or recurrent follicular tumors with bland cytomorphological features (Fig. 4C and D). In rare cases, it can aid in identifying collision tumors composed of RAS-mutated follicular tumors and other RAS-negative tumors (Fig. 4E and F) [13].

Utility of RAS Q61R IHC in diagnosing follicular neoplasms. A A RAS-mutant FA showing diffuse moderate cytoplasmic staining with RAS IHC (400 ×). B A RAS-mutant widely invasive FTC with a distinctive membranous staining pattern with RAS IHC (400 ×). C An encapsulated follicular neoplasm with a suspected invasive satellite nodule (H&E, 10 ×). D RAS IHC highlights the tumor and its satellite nodule, confirming the diagnosis of minimally invasive FTC (10 ×). E A nodule with morphologic features of follicular adenoma (asterisk) immediately adjacent to a “hyperplastic-appearing” nodule (H&E, 50 ×). F RAS IHC is positive in the smaller “hyperplastic-appearing” nodule and negative in the follicular adenoma, confirming that both are molecularly unrelated and excluding the possibility of an FTC with an invasive satellite nodule (200 ×)

Utility of RAS Q61R IHC in diagnosing follicular neoplasms. A Widely invasive follicular thyroid carcinoma with a multinodular goiter-like appearance (H&E, 2.5 ×). B RAS IHC highlights the infiltrative nodules confirming the diagnosis (10 ×). C Cytoarchitecturally bland goiter-like metastasis of an FTC in a needle biopsy of the lung (H&E, 200 ×). D Staining with RAS IHC confirms the neoplastic nature of this metastatic lesion (200 ×). E Collision of a RET::NCOA4 PTC on the left with a RAS-mutant follicular neoplasm on the right (H&E, 100 ×). F RAS IHC highlights the interface of the molecularly discordant tumors (100 ×)

In sporadic medullary thyroid carcinomas (MTC), in which the majority harbor RET mutations, up to 20% can have mutually exclusive alterations in the RAS family of genes, with the most common being an HRAS p.Q61R mutation [36]. In a recent tri-institutional cohort study of MTCs, RAS Q61R IHC was demonstrated to have exceptional sensitivity (100%), specificity (100%), positive predictive value (100%), and negative predictive value (100%) in detecting the RAS p.Q61R mutation in both cytologic and histologic samples of sporadic MTCs when staining was carefully ascertained to be at least weak (staining score of > 1) and/or with a membranous accentuation [37]. Membranous staining with the RAS Q61R IHC was found to be 100% predictive of RET-negative germline testing, highlighting its potential role as an inexpensive and rapid modality to screen patients who will undergo RET germline testing for MTC.

NTRK gene rearrangements are found in 2.3–3.4% of PTC, with the most common gene fusion being ETV6::NTRK3 [1, 38]. Our previous studies found the specificity and sensitivity of pan-TRK IHC (clone EPR 17341) to be 100% and 41.7% for detecting NTRK 1/3 rearrangements in BRAF V600E-negative PTCs in tissue microarray, respectively [39]. Based on a pooled analysis, pan-TRK IHC has shown lower sensitivity in detecting NTRK3 rearrangements (65.8%) than NTRK1 rearrangements (87.5%) [40]. Pan-TRK IHC has varied staining patterns in NTRK-rearranged PTCs. ETV6::NTRK3 fusion-associated PTCs usually showed focal and limited staining in the cytoplasm and nucleus with variable staining intensity, while non-ETV6::NTRK3 cases frequently showed a more homogeneous and diffuse staining pattern, usually in the cytoplasm with or without membranous attenuation [39]. Interestingly, the heterogeneous staining pattern in ETV6::NTRK3 fusion-associated PTCs tends to be stronger in the tumor periphery and becomes weaker to subdiagnostic towards the center, a phenomenon that might be dismissed as artifactual in nature and considered to be non-diagnostic [41]. A knowledge of this heterogeneous staining pattern should likewise alert pathologists to the possibility of a false-negative result when interpreting pan-TRK IHC in limited biopsy samples. Similar to other kinase fusion-associated tumors, NTRK-rearranged PTCs tend to display non-classical histology enriched with microfollicles, multinodular permeative growth, dense intratumoral sclerosis, and subtle nuclear features, which may pose a diagnostic challenge [39, 42]. Thus, staining with pan-TRK IHC is a low-cost method to identify these cases that would require further molecular confirmation (Fig. 5A to C).

Utility of pan-TRK and ALK IHC in diagnosing thyroid carcinomas. A NTRK-rearranged PTC shows a multinodular permeative growth (H&E, 50 ×). B Pan-TRK IHC highlights the tumor’s multinodular pattern of invasion (50 ×). C Nuclear and cytoplasmic expression of pan-TRK IHC consistent with ETV6::NTRK3 fusion (200 ×). D Low power view of a well-demarcated ALK-rearranged PTC with architectural heterogeneity (H&E, 10 ×). E High power view demonstrates the ALK-rearranged PTC with the left area showing classic PTC nuclear features juxtaposed to the right area resembling follicular nodular disease with bland cytologic features (H&E, 200 ×). F ALK IHC shows strong diffuse expression in the tumor (200 ×)

ALK gene rearrangements are found in 1.2% of PTCs, and the most frequent are STRN::ALK (45.8%) and EML4::ALK (31.4%). The antibody clone 5A4 has shown excellent sensitivity (100%) and modest specificity (75%) in detecting the ALK gene rearrangement in PTCs, seen as diffuse expression in the cytoplasm of tumor cells [43]. Additionally, we have used the D5F3 clone to identify eight cases of ALK-rearranged PTCs in a recent study, and six of them were confirmed through targeted RNA next-generation sequencing (NGS) (specificity 100%) [44]. ALK gene fusion-associated thyroid carcinomas can show subtle cytologic nuclear features resembling RAS-like tumors and present with a myriad of architectural patterns such as classic, classic with a predominant microfollicular growth, infiltrative follicular, and solid-trabecular [43, 44]. Rarely, a tumor can be unusually well-demarcated and have cytologic heterogeneity, with cells displaying robust PTC nuclear changes mixed with those displaying only subtle or RAS-like nuclear features (Fig. 5D to F). Thus, the identification of ALK-rearranged PTC can be diagnostically challenging without the aid of the ALK IHC.

ATC is the most lethal form of thyroid cancer, with a worldwide incidence of 1–4% and a dismal survival rate [14]. ATC may be encountered as a de novo tumor or may be accompanied by a well-differentiated or poorly differentiated thyroid carcinoma. ATC harbors BRAF p.V600E (25.9–45.0%) and RAS mutations (24.0–40.7%) as early molecular events, while TP53 (40–80%) and TERT promoter mutations (30–75%) are additional late molecular events and the most frequent genetic changes seen in this tumor [45, 46]. The BRAF p.V600E mutation is significantly associated with ATC exhibiting a squamous cell carcinoma phenotype and/or co-existing PTC. Next-generation IHC for BRAF V600E and RAS Q61R have been employed to identify the molecular drivers of ATC, showing diffuse weak to moderate expression in the tumor cytoplasm (Fig. 6A to B). When ATC co-exists with a well-differentiated thyroid carcinoma, the staining is typically preserved in both components, consistent with tumor high-grade transformation (Fig. 6C to D). However, the expression may be diffusely attenuated in the ATC component, particularly in cases with predominantly sarcomatous differentiati