Abstract
Purpose
Primary salivary gland tumors form a heterogeneous group with diagnostic challenges. Sex-determining region Y-related homeobox 10 (SOX10) is a transcription factor studied in primary salivary gland tumors for over a decade. To date, no systematic review or meta-analysis has been conducted on its expression in salivary gland tumors.
Methods
A PubMed literature search identified all studies in which SOX10 was evaluated by immunohistochemistry in primary salivary gland tumors. Information on SOX10 positivity, staining pattern, percentage of positive cells, and information on the antibody source and clone were tabulated in Microsoft Excel. The meta-analysis was performed with OpenMeta Analyst. The meta-analysis adheres to PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines.
Results
The meta-analysis included 35 studies with 1146 histological and 35 cytological samples immunostained with SOX10. The list of SOX10-positive tumors according to our meta-analysis includes pleomorphic adenoma (pooled meta-analytic estimate of SOX10 positivity 95.8%, 95% CI 91.7–99.9%), polymorphous adenocarcinoma (pooled estimate 97.4%, 95% CI 92.3–102.6%), canalicular adenoma (pooled estimate 98.9%, 95% CI 96.4–101.4%), basal cell adenoma (pooled estimate 91.3%, 95% CI 80.2–102.5%), adenoid cystic carcinoma (pooled estimate 96.6%, 95% CI 94.1–99.2%), myoepithelioma (pooled estimate 75.2%, 95% CI 52.0–98.4%), acinic cell carcinoma (pooled estimate 97.4%, 95% CI 94.1–100.6%), intercalated duct type of intraductal carcinoma (pooled estimate 90.8%, 95% CI 75.9–105.8%), and microsecretory carcinoma (pooled estimate 96.1%, 95% CI 87.2–105.0%). Limited evidence suggests that basal cell adenocarcinoma, epithelial–myoepithelial carcinoma, and myoepithelial carcinoma are also SOX10 positive. SOX10-negative entities include Warthin’s tumor (pooled estimate 2.3%, 95% CI 0–7.0%), apocrine type of intraductal carcinoma (pooled estimate 10.3%, 95% CI − 10.4–31.1%), salivary duct carcinoma (pooled estimate 3.7%, 95% CI − 0.6–8.1%), and oncocytoma (pooled estimate 8.0%, 95% CI − 2.7–18.7%). Polyclonal antibodies and clones EP268, N-20, and A-2 provided similar results. Clone BC34 stained a lower percentage of SOX10-positive tumors.
Conclusion
This meta-analysis lists SOX10-positive and -negative tumors. This demonstrates the necessity of subtype-specific reporting and stating the antibody clone used in immunohistochemical studies.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1007/s12105-026-01975-3.
Keywords: SOX10, Salivary gland, Meta-analysis, Immunohistochemistry
Introduction
Primary salivary gland tumors are neoplastic diseases that originate from both the major and minor salivary glands. They are a heterogeneous group, with the World Health Organization´s (WHO) Classification of Tumors recognizing 15 benign and 21 malignant tumors [1]. Based on the histological similarities among the entities, the tumors can be classified as basaloid salivary gland tumors, tumors with oncocytic features, and tumors with clear cell features. However, several entities do not belong to any of these groups. This variety poses a diagnostic challenge and makes immunohistochemical studies vital for differentiating salivary gland tumors [2].
Sex-determining region Y (SRY)-related homeobox 10 (SOX10) is a transcription factor involved in the development of the neural crest, cartilage, glial cells, and sex determination. Therefore, it has been used as an immunohistochemical marker for tumors arising from cells with a neural crest origin, such as schwannomas and melanomas [3]. In normal tissues, nuclear SOX10 positivity is encountered in the myoepithelial cells of the mammary, salivary, and bronchial glands. In addition, as noted by Koutlas et al. [4], the positive nuclear staining of SOX10 has been recorded in the acinar, luminal, and abluminal cells of salivary glands.
Although SOX10 positivity in salivary gland tumors has been the subject of constant research for more than a decade, the exact range of SOX10-positive entities remains unclear. For instance, only four tumors are explicitly stated as SOX10 positive in both the WHO Classification of Tumors [1] and the Diagnostic Immunohistochemistry–Theranostic and Genomic Applications book [5]: pleomorphic adenoma, myoepithelioma, secretory carcinoma, and intraductal carcinoma. Nevertheless, each book lists another eight [1] and four [5] SOX10-positive tumors, respectively. While entities where SOX10 was not addressed were not necessarily considered SOX10 negative, the difference may demonstrate a differing view of the value of SOX10 for the salivary gland entities. Furthermore, a study with a large selection of entities conducted by Hsieh et al. [6] not only supports the claims of both books but also expands the list of SOX10-positive tumors.
Therefore, in this meta-analysis, a literature search found all original studies in which primary salivary gland tumors were immunostained with the SOX10 antibody. The objective of this study was to determine which primary salivary gland tumors are SOX10 positive and how reliable SOX10 is as a diagnostic marker in differential diagnostic work-ups.
Materials and Methods
Search Criteria
Originally, the objective of this study was to determine SOX10 expression in the cytological specimens of primary salivary gland tumors. A PubMed literature search was conducted between June 15, 2025, and August 4, 2025. The search keywords used were “SOX10 and salivary gland and (FNA or cytology).” To avoid scenarios in which “salivary gland” was not used in the study, additional search words were used. These included the three major salivary glands (parotid, submandibular, and sublingual glands) and all primary salivary gland tumors described as SOX10 positive in either the WHO Classification of Tumors [1] or Diagnostic Immunohistochemistry [5]. The complete list of search terms used is illustrated in Suppl. Table 1. This search yielded 123 articles, one of which fulfilled the criterion for the cytological material studied.
Because the number of cases was insufficient for a meta-analysis, it was decided to expand the goal of this study to include both cytological and histological samples. A PubMed literature search was performed between August 25, 2025, and September 1, 2025, with the search terms “SOX10 and salivary gland.” Since the use of extra search terms in place of “salivary gland” provided no additional benefit when searching for studies on cytological samples, it was concluded that “SOX10 and salivary gland” is sufficient.
Study Inclusion and Exclusion Criteria
The following criteria were used to include studies in the title and abstract screening: (1) The research includes primary salivary gland tumors, and (2) SOX10 positivity has been studied by immunohistochemistry. Meta-analyses, reviews, case reports, and non-English studies were excluded. However, if a literature review accompanied an original study, the article was not excluded, and only the data from the original study were considered. In a full text reading, studies were excluded if the method used to determine SOX10 positivity was ambiguous or not described. Studies were also excluded if the number of SOX10-positive tumors was not stated or if discrepancies regarding the number of SOX10-positive tumors were detected, for example, between the text and tables (Fig. 1).
Fig. 1.
Flow diagram describing the process of identifying studies on SOX10 positivity in salivary gland tumors. *For original studies accompanied by a literature review, only the materials of the original study were considered
Data Collection
The following data were gathered on a Microsoft Excel sheet (Version 2512, build 16.0.19530.20184, 64-bit): PubMed ID, the first author, year of publication, country of origin, continent, whether the study was institutional, multi-institutional, or international, period of study in years, the type of hospital or laboratory where the study was conducted, and the total number of samples studied. The second Excel sheet contained information on the tumors, themselves: histological and/or cytological diagnosis, tumor grade or subtype, if applicable, the total number of tumors studied and immunostained with SOX10 and the number of SOX10-positive tumors, type of SOX10-positive cells, SOX10 staining pattern, SOX10 staining intensity, percentage of SOX10-positive cells, topography of the tumor and the source, a clone, and the dilution of the SOX10 antibody used. The second Excel sheet was employed for histological samples, and the third Excel sheet was created for cytological samples. The information collected was the same as in the second Excel sheet.
Meta-Analysis
Statistical analyses were performed using OpenMeta [Analyst] Version 0.24.1. For each tumor entity, the proportion of SOX10-positive cases was pooled across studies using a random-effects model with the DerSimonian–Laird estimator. Quantitative syntheses were performed only for entities evaluated in at least two independent studies. Pooled proportions and corresponding 95% confidence intervals were presented in forest plots. For each tumor entity, two different measures of SOX10 positivity were distinguished. The crude (observed) proportion was calculated as the total number of SOX10-positive cases divided by the total number of evaluated cases across all included studies. In contrast, the pooled proportion represents a meta-analytic estimate obtained by combining study-specific proportions using the random-effects model described above. Therefore, the pooled estimate is not mathematically equivalent to the crude proportion and may differ from the percentage obtained by simply aggregating all positive and negative cases. Throughout the manuscript, the terms “crude/observed proportion” and “pooled meta-analytic estimate” are used to distinguish these measures. Because the confidence intervals are model-based statistical estimates, their limits may occasionally extend slightly beyond the theoretical range of 0–100%. Such values should not be interpreted as biologically possible proportions but as a consequence of the statistical estimation procedure. Between-study heterogeneity was assessed using Cochran’s Q test and quantified using the I2 inconsistency statistic. Entities represented by a single study were summarized descriptively. Histological subtypes were analyzed separately when sufficient data were available. The meta-analysis was reported as per PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) 2020 Guidelines [7].
Results
Study Cohort
The literature search yielded 155 articles, of which 40 were selected for a full-text reading based on the inclusion and exclusion criteria. Five studies were excluded after a full-text reading. Of the 35 studies included in the meta-analysis, 14 originated from the USA [8–21], five from Japan [22–26], three from Taiwan [6, 27, 28], three from the Czech Republic [29–31], two from Germany [32, 33], and one each from Brazil [34], Canada [35], Iran [36], Mexico [37], the People’s Republic of China [38], the Republic of Korea [39], Singapore [40], and South Africa [41]. The studies were published between 2013 and 2025. According to the study data periods stated in 22 studies, the series represented at least 343 study years. The number of histological samples across the studies was 1280, of which 1146 were immunostained with SOX10. In total, 35 cytologic samples were also included, all immunostained with SOX10. A flow diagram of the study identification process is presented in Fig. 1, and an overview of the studies is illustrated in Suppl. Table 2.
Topography of Tumors
All major salivary glands are represented in this meta-analysis, with 226 parotid, 15 submandibular, and two sublingual gland tumors. In total, 165 tumors originated from minor salivary glands (MSG), and the exact location was stated of 98: 25 in the hard palate, 21 in the buccal mucosa, seven in the labial mucosa, six in the palate, five in the upper lip, four in the tongue, three in the lip, three in the soft palate, two in the maxilla, four in the oropharynx, two in the retromolar trigone, and one each in the alveolar ridge, the angle of the mandible, the base of the tongue, the bronchus, the floor of the mouth, the gingiva, the junction of the hard and soft palate, the labial vestibule, the lower lip, the mandible, the nasopharynx, the maxillary sinus, the sphenoid sinus, and the parapharyngeal, the parabronchial, and the pharyngeal compartments. Sixty-seven MSG tumors were in the upper lip, buccal mucosa, and palate. However, the exact number of samples in each location was not stated. Similarly, 39 tumors were in either the parotid or submandibular gland. In addition, 22 tumors were in the parotid, sublingual, and submandibular glands or minor salivary glands in the tongue, the base of the tongue, or the palate. Moreover, 16 tumors were in the parotid gland or lip, accessory parotid, or pharynx MSGs. The location was not reported in 830 (70%) samples.
Basaloid Salivary Gland Tumors
Basaloid salivary gland tumors include pleomorphic adenoma (PA), basal cell adenoma (BCA), basal cell adenocarcinoma (BCAC), adenoid cystic carcinoma (AdCC), myoepithelial carcinoma, epithelial–myoepithelial carcinoma (EMC), and polymorphous adenocarcinoma (PAC). Table 1 presents the results of this group.
Table 1.
Summary of SOX10 expression and diagnostic applications in basaloid salivary gland tumors
| Tumor type | Studies (n) | SOX10 positive/total (n) | Crude SOX10 positivity (%) | Pooled estimate of SOX10 positivity (95% CI) | Heterogeneity I2 | P-value | Diagnostic implication |
|---|---|---|---|---|---|---|---|
| Pleomorphic adenoma (PA) | 7 | 163/174 | 93.7 | 95.8% (91.7–99.9) | 52.4% | 0.050 | Excellent marker |
| Basal cell adenoma (BCA) | 4 | 20/21 | 95.2 | 91.3% (80.2–102.5) | 0% | 0.906 | Very strong expression |
| Basal cell adenocarcinoma (BCAC) | 2 | 13/15 | 86.6 | 84.8% (67.3–102.3) | 0% | 0.738 | Strong but limited data |
| Adenoid cystic carcinoma (AdCC) | 9 | 176/181 | 97.2 | 96.6% (94.1–99.2) | 0% | 0.982 | Very high and consistent expression |
| Myoepithelioma | 3 | 10/14 | 71.4 | 75.2% (52.0–98.4) | 23.5% | 0.271 | High expression |
| Myoepithelial carcinoma | 2 | 2/2 | 100 | 75.0% (15.0–100.0) | 0% | 0.999 | Very small sample |
| Epithelial–myoepithelial carcinoma (EMC) | 3 | 9/9 | 100 | 89.0% (71.6–106.5) | 0% | 0.872 | Strong positivity (limited data) |
| Polymorphous adenocarcinoma (PAC) | 3 | 32/33 | 97.0 | 97.4% (92.3–102.6) | 0% | 0.502 | Excellent marker |
Seven studies reported PA SOX10 immunohistochemistry, with 174 samples studied. Of these, 163 were SOX10 positive, corresponding to a crude observed positivity rate of 93.7%. The pooled meta-analytic estimate of SOX10 positivity was 95.8% (Suppl. Figure 1a). Correspondingly, SOX10 is an excellent marker of PA.
Four studies featured BCAs, and 21 tumors were studied. Of these, 20 were SOX10 positive, corresponding to a crude observed positivity rate of 95.2% and a pooled meta-analytic estimate of SOX10 positivity of 91.3% (Suppl. Figure 1b). SOX10 expression was very strong in the BCA.
Only two series studied BCACs. Thirteen of the 15 tumors studied were SOX10 positive, corresponding to a crude observed positivity rate of 86.7%. The pooled meta-analytic estimate of SOX10 positivity was 84.8% (Suppl. Figure 1c). SOX10 expression can be considered very strong, though it is based on limited data in BCAC.
AdCC was a focus in nine studies, with 181 immunohistochemistry results available. SOX10 was positive in 176 samples, corresponding to a crude observed positivity rate of 97.2%. The pooled meta-analytic estimate of SOX10 positivity was 96.6% (Suppl. Figure 1d). SOX10 expression in AdCC is high and consistent.
Three studies featured myoepithelioma, with 14 tumors studied. Of these, 10 were SOX10 positive, with a crude observed positivity rate of 71.4% and a pooled meta-analytic estimate of SOX10 positivity being 75.2% (Suppl. Figure 1e). Accordingly, SOX10 expression in myoepithelioma was high.
Myoepithelial carcinomas were covered in two studies, with two tumors studied overall. Both were SOX10 positive, corresponding to a crude observed positivity rate of 100%. However, the pooled meta-analytic estimate of SOX10 positivity was 75% (Suppl. Figure 1f). The pooled estimate reflects the meta-analytic model and should not be interpreted as the crude observed proportion. Although both tumors were SOX10 positive and heterogeneity was low, the findings were based on a very small sample size.
EMC was featured in three studies, and nine tumors were studied overall. All were SOX10 positive, with a crude observed positivity rate of 100% and a pooled meta-analytic estimate of SOX10 positivity of 89% (Suppl. Figure 1g). Similarly to myoepithelial carcinoma, the difference between the crude observed positivity rate and meta-analytic estimate of SOX10 positivity reflects the meta-analytic model. SOX10 positivity in EMC is strong but based on limited data.
PAC was studied in three articles comprising 33 tumors. In total, 32 were SOX10 positive, corresponding to a crude observed positivity rate of 97.0% and the pooled meta-analytic estimate of SOX10 positivity was 97.4% (Suppl. Figure 1h). Hence, SOX10 is an excellent marker of PAC.
Salivary Gland Tumors with Oncocytic Features
Warthin’s tumor (WT), oncocytoma, acinic cell carcinoma (AciCC), secretory carcinoma (SC), mucoepidermoid carcinoma (MEC), and salivary duct carcinoma (SDC) are considered salivary gland tumors with oncocytic features. Table 2 presents the results of this group.
Table 2.
Summary of SOX10 expression and diagnostic application in salivary gland tumors with oncocytic features
| Tumor type | Studies (n) | SOX10 positive/total (n) | Crude SOX10 positivity (%) | Pooled estimate of SOX10 positivity (95% CI) | Heterogeneity I2 | P-value | Diagnostic implication |
|---|---|---|---|---|---|---|---|
| Warthin tumor (WT) | 3 | 0/36 | 0 | 2.3% (0–7.0) | 0% | 0.704 | Typically negative → strong differential value |
| Oncocytoma | 5 | 0/19 | 0 |
8.0% (− 2.7–18.7) |
0% | 0.944 | Typically negative |
| Acinic cell carcinoma (AciCC) | 7 | 92/95 | 96.8 | 97.4% (94.1–100.6) | 7.7% | 0.370 | Reliable, strong expression |
| Secretory carcinoma (SC) | 9 | 57/68 | 83.8 | 79.9% (55.1–104.7) | 95.6% | <0.001 | Considerable variability |
| Mucoepidermoid carcinoma (MEC) | 9 | 28/175 | 16.0 | 20.2% (2.8–37.6) | 95.1% | <0.001 | Mostly negative |
| Salivary duct carcinoma (SDC) | 8 | 0/63 | 0 |
3.7% (− 0.6–8.1) |
0% | 0.981 | Typically negative |
WT was featured in three studies, with a total of 36 tumors. None was SOX10 positive, corresponding to a crude observed positivity rate of 0% and making the pooled meta-analytic estimate of SOX10 positivity 2.3% (Suppl. Figure 2a). SOX10 can be considered typically negative in WT, giving it a high value in the differential diagnostics of salivary gland tumors with oncocytic features.
Oncocytomas were featured in only five studies, with a total of 19 tumors. None was SOX10 positive, corresponding to a crude observed positivity rate of 0%. The pooled meta-analytic estimate of SOX10 positivity in oncocytoma was 8% (Suppl. Figure 2b). The pooled estimate reflects the meta-analytic model and should not be interpreted as the crude observed proportion. Consequently, oncocytoma is typically negative in SOX10 immunohistochemistry.
AciCC was targeted in seven studies, with 95 tumors in total. Of these, 92 tumors were SOX10 positive, corresponding to a crude observed positivity rate of 96.8% and a pooled meta-analytic estimate of SOX10 positivity of 97.4% (Suppl. Figure 2c). Therefore, SOX10 expression in AciCC is reliable and strong.
SC was investigated in nine studies with 68 tumors in total. Of these, 57 tumors were SOX10 positive, corresponding to a crude observed positivity rate of 83.8%. The pooled meta-analytic estimate of SOX10 positivity was 79.9% (Suppl. Figure 2d). SC demonstrated considerable variability in SOX10 staining. No histological subtype or grade was stated, leaving the reason for the variability unclear.
MEC was featured in nine studies with 175 tumors in total. Of these, 28 tumors were SOX10 positive, corresponding to a crude observed positivity rate of 16.0%. The pooled meta-analytic estimate of SOX10 positivity was 20.2% (Suppl. Figure 2e). SOX10 was mostly negative in MEC.
SDC was the target of eight studies, with 63 tumors in total. None was SOX10 positive, corresponding to a crude observed positivity rate of 0%. The pooled meta-analytic estimate of SOX10 positivity was 3.7% (Suppl. Figure 2f). Typically, SDC can be considered SOX10 negative.
Salivary Gland Tumors with Clear Cell Features
Myoepitheliomas, myoepithelial carcinomas, AciCCs, MECs, and hyalinizing clear cell carcinomas (HCCCs) are salivary gland tumors with clear cell features. Due to morphological heterogeneity, the following entities can be classified as salivary gland tumors with basaloid or oncocytic features: myoepitheliomas, myoepithelial carcinomas, AciCCs, and MECs. The results of these tumors were described earlier with salivary gland tumors having basaloid or oncocytic features. Table 3 summarizes the results of salivary gland tumors with clear cell features.
Table 3.
Summary of SOX10 expression and diagnostic application in salivary gland tumors with clear cell features
| Tumor type | Studies (n) | SOX10 positive/total (n) | Crude SOX10 positivity (%) | Pooled estimate of SOX10 positivity (95% CI) | Heterogeneity I2 | P-value | Diagnostic implication |
|---|---|---|---|---|---|---|---|
| Myoepithelioma | 3 | 10/14 | 71.4 | 75.2% (52.0–98.4) | 23.5% | 0.271 | High expression |
| Myoepithelial carcinoma | 2 | 2/2 | 100 | 75.0% (15.0–100.0) | 0% | 0.999 | Very small sample |
| Acinic cell carcinoma (AciCC) | 7 | 92/95 | 96.8 | 97.4% (94.1–100.6) | 7.7% | 0.370 | Reliable, strong expression |
| Mucoepidermoid carcinoma (MEC) | 9 | 28/175 | 16.0 | 20.2% (2.8–37.6) | 95.1% | <0.001 | Mostly negative |
| Hyalinizing clear cell carcinoma (HCCC) | 3 | 2/24 | 8.3 |
8.8% (− 1.9–19.6) |
0% | 0.641 | Mostly negative |
HCCCs were featured in three studies with 24 tumors in total. Two were SOX10 positive, corresponding to a crude observed positivity rate of 8.3% and making the pooled meta-analytic estimate of SOX10 positivity 8.8% (Suppl. Figure 3). With these findings, HCCC can be considered mostly SOX10 negative.
Other Analyzed Entities
Adenocarcinomas not otherwise specified (NOS) were researched in four studies, with nine tumors in total. Four were SOX10 positive, corresponding to a crude observed positivity rate of 44.4% and a pooled meta-analytic estimate of SOX10 positivity of 39.8% (Suppl. Figure 4a). Due to the variable results, the utility of SOX10 in the diagnostics of adenocarcinoma NOS is limited. Table 4 summarizes the other analyzed entities.
Table 4.
Summary of SOX10 expression and diagnostic application in other analyzed entities
| Tumor type | Studies (n) | SOX10 positive/total (n) | Crude SOX10 positivity (n, %) | Pooled estimate of SOX10 positivity (95% CI) | Heterogeneity I2 | P-value | Diagnostic implication |
|---|---|---|---|---|---|---|---|
| Adenocarcinoma, NOS | 4 | 4/9 | 44.4 |
39.8% (− 4.3–84.0) |
78.7% | 0.003 | Very variable; limited utility |
| Canalicular adenoma (CA) | 2 | 62/62 | 100 | 98.9% (96.4–101.4) | 0% | 0.575 | Essentially universally positive |
| Intraductal carcinoma– combined | 5 | 15/29 | 51.7 | 56.4% (17.6–95.1) | 88.2% | <0.001 | Strong heterogeneity → subtype-specific reporting recommended |
| Intraductal carcinoma– intercalated type | 3 | 11/11 | 100 | 90.8% (75.9–105.8) | 0% | 0.854 | Strong positivity |
| Intraductal carcinoma– apocrine type | 2 | 0/6 | 0 |
10.3% (− 10.4–31.1) |
0% | 0.610 | Typically negative |
| Microsecretory carcinoma | 2 | 16/16 | 100 | 96.1% (87.2–105.0) | 0% | 0.545 | Very high expression |
| Sialadenoma papilliferum (SP) (all subtypes) | 2 | 10/20 | 50.0 |
50.0% (0–100) |
0% | 1.000 | Insufficient evidence |
Canalicular adenomas (CAs) were the focus of two studies with 62 tumors in total. All were SOX10 positive, corresponding to a crude observed positivity rate of 100%, and the pooled meta-analytic estimate of SOX10 positivity was 98.9% (Suppl. Figure 4b). As such, SOX10 is essentially universally positive in CA.
All types of intraductal carcinoma were featured in five studies, with 29 tumors in total. Fifteen were SOX10 positive, corresponding to a crude observed positivity rate of 51.7%. The pooled meta-analytic estimate of SOX10 positivity was 56.4% (Suppl. Figure 4c). Two subtypes of intraductal carcinoma had sufficient data for further analysis. The first was the intercalated duct type, which was examined in three studies with 11 tumors in total. All were SOX10 positive, corresponding to a crude observed positivity rate of 100% and a pooled meta-analytic estimate of SOX10 positivity of 90.8% (Suppl. Figure 4d). The other subtype analyzed was the apocrine type, which was evaluated by two studies with six tumors in total. None was SOX10 positive, corresponding to a crude observed positivity rate of 0%. The pooled meta-analytic estimate of SOX10 positivity was 10.3% (Suppl. Figure 4e). In conclusion, SOX10 is a strong marker of the intercalated subtype of intraductal carcinomas and is mostly negative in the apocrine subtype. These findings demonstrate the necessity of subtype-specific reporting in intraductal carcinomas.
Microsecretory carcinomas were featured in two studies, with 16 tumors in total. All were SOX10 positive, corresponding to a crude observed positivity rate of 100%, and the pooled meta-analytic estimate of SOX10 positivity was 96.1% (Suppl. Figure 4f). These findings imply that the expression of SOX10 in microsecretory carcinomas is very high.
Sialadenoma papilliferum (SP) was analyzed regarding all the subtypes combined. Two studies assessed them, with 20 tumors in total. Ten were SOX10 positive, corresponding to a crude observed positivity rate of 50%, and the pooled meta-analytic estimate of SOX10 positivity was 50% (Suppl. Figure 4g). The evidence was insufficient to draw any conclusions of SOX10 positivity regarding SP when analyzed with all subtypes combined.
Entities not Analyzed
Meta-analyses of SOX10 positivity were not conducted on tumors that were not featured in at least two independent studies. Therefore, the evaluation of SOX10 positivity in these entities is based on very limited evidence, and their SOX10 positivity is summarized only in Suppl. Table 3.
Cytological Samples
Only Schmitt et al. [10] evaluated cytological samples. The material used was a cell block. The findings of this study have been analyzed with histologic samples in forest plot construction. Eight of 13 samples of AciCC were SOX10 positive. They were stained in a nuclear pattern, and the percentage of cells stained was 50%. WT, MEC, and oncocytoma were SOX10 negative (0/16, 0/3, and 0/3, respectively). Polyclonal antibody was used for all cytological samples.
SOX10 Staining Pattern, Cell Types, and Intensity
In all cases where the SOX10 staining pattern was explicitly stated, it was nuclear. A nuclear SOX10 staining pattern was reported in 29 studies (83%). No membranous or cytoplasmic staining was reported.
Abluminal cell positivity was reported in 82 samples in total, of which 61 were AdCC, 13 were canalicular adenoma-like PA, and eight were BCA [14, 31]. Luminal and abluminal cell positivity was found in 22 samples of AdCC, five samples of EMC, and three samples of AciCC [22, 23]. Luminal epithelial cell positivity was reported in 21 samples of sclerosing polycystic adenoma [30]. Canalicular cell positivity was reported in 16 samples of canalicular-like PA [33]. Endophytic ductal cell positivity was reported in 10 samples of SP [19, 27]. Chondroid, epitheloid, spindle, and plasmacytoid cell positivity was found in 10 samples of PA. However, the exact number of each cell type stained was not stated [22]. Epithelial and mesenchymal cell positivity was reported in nine samples of PA, and in the same study, epithelial and myoepithelial positivity was reported in three samples of EMC [6]. Acinar cell positivity was found in nine samples of the mucoacinar subtype of MEC [40]. Polygonal cell positivity was found in six samples of MEC [6]. Spindle cell positivity was reported in four samples of myoepithelioma and one sample of myoepithelial carcinoma [22]. Intercalated duct cell positivity was found in two, and a combination of intercalated duct and oncocytic cell positivity was noticed in three samples of intraductal carcinoma [16]. SOX10 positivity in serous acini and mucous acini was reported in two HCCC samples [26]. In most studies of the meta-analysis, the text did not explicitly state the exact cell type stained positive for SOX10.
Information on SOX10 staining intensity and the SOX10 staining percentage in the analyzed publications was equally rare. High staining intensity was reported in PA [22], SP [19, 27], hybrid and intercalated duct types of intraductal carcinoma [16], EMC [22], SC [11, 36], AdCC [9], PAC [13], MEC [29], striated duct adenoma [18], AciCC [22], CA [8], microsecretory adenocarcinoma [15], myoepithelial carcinoma, and myoepithelioma [22]. A variable, high to low intensity, was reported in AdCC [12, 13, 22]. Intermediate staining intensity was reported in hybrid intraductal carcinoma [16], while low staining intensity was reported in HCCC [26]. SOX10 staining percentages were rarely reported as exact percentages, and the range of percentages in SOX10 positive cases ranged from 10%–100%.
SOX10 Antibodies Used
In terms of antibody source and dilution, nine studies used Cell Marque polyclonal antibody (Rocklin, CA, USA) [10, 12, 23, 29, 36] or the following clones, if stated: EP268 [11], SP267 [30], and Catalog #PA0813 [13]. Four studies used the Ventana Medical Systems (Tucson, AZ, USA) antibody, which was ready to use. The clone used was SP267 [31] or was not stated [15, 17, 20]. Four studies used Bio SB (Santa Barbara, CA, USA) antibody, which was ready to use, and all used clone EP268 [6, 27, 28, 39]. Four studies used Santa Cruz Biotechnology’s (Santa Cruz, CA, USA) antibody with 1:200 and 1:500 dilutions, and the clones used were N-20 [8, 22, 26] and A-2 (sc-365692) [25]. Two studies used Abcam (Cambridge, UK) antibody with the dilutions 1:100 [34] and 1:25 [38], with the former using the clone SP267. Two studies used DCS polyclonal antibodies with a 1:25 dilution [32, 33]. Two studies used the Biocare Medica predilute antibody (clone BC34) [19, 40]. One study used Leica Biosystems (CA, USA) with the clone BC34 and an unknown dilution [14]. One study used Diagnostic BioSystems’ (Pleasanton, CA, USA) antibody with 1:100 dilution and the clone EP268 [24]. One study used a Biocare antibody with a 1:100 dilution and the clone BC34 [16]. The source, clone, and dilution were not stated in five (14%) studies [18, 21, 35, 37, 41]. The different antibodies used are tabulated in Suppl. Table 4.
Discussion
In this meta-analysis of 35 studies, 32 histological entities and 1146 histological samples in total, along with 35 cytological samples, were evaluated for SOX10 immunoexpression. SOX10 showed a pooled meta-analytic positivity estimate of 95.8% in PA, making it an excellent marker for PA. PA was explicitly stated as SOX10 positive by both the WHO Classification of Tumors and Diagnostic Immunohistochemistry [1, 5]. In this meta-analysis, five studies reported 100% SOX10 positivity in PA [6, 21, 22, 31, 39]. In contrast, the lowest PA SOX10 positivity (76%) was found in Ko et al. [14]. Only this study evaluated SOX10 positivity in PA with the clone BC34. These findings entrench PA as a SOX10-positive tumor.
PAC is another entity in which SOX10 is an excellent marker (pooled estimate of SOX10 positivity 97.4%, 95% CI 92.3–102.6%). 95% confidence interval exceeding 100% reflects the meta-analytic model used and should not be interpreted as a biologically possible result. Unlike PA, PAC was stated as SOX10 positive by Diagnostic Immunohistochemistry [5], while WHO Classification of Tumors did not address the SOX10 positivity of PAC. Of the three studies in this meta-analysis, two reported total SOX10 positivity [13, 39]. Conversely, one study had only a single sample of SOX10-negative PAC [14]. As was the case with PA, the SOX10 clone used in the study was BC34. The other studies used EP268 [39] and Cell Marque (Rocklin, CA, USA) Catalog #PA0813 [13].
CA can be considered essentially universally SOX10 positive, with a crude observed positivity rate of 100%. The pooled meta-analytic estimate of SOX10 positivity in CA was 98.9% (95% CI 96.4–101.4%). As a result, there was no disparity in SOX10 positivity between the SOX10 clones.
Several entities had strong and consistent SOX10 immunoexpression. BCA (pooled estimate 91.3%, 95% CI 80.2–102.5%) had total SOX10 positivity in two studies using the EP268 clone [6, 39] and in one study using an unknown clone [21]. In contrast, eight out of nine samples of BCA were SOX10 positive with the BC34 clone [14]. The result should be interpreted with caution given that the sample size was small. AdCC also had strong SOX10 expression, with a pooled estimate of SOX10 positivity of 96.6% (95% CI 94.1–99.2%). Unlike PA, PAC, and BCA, there was no distinction between BC34 and the other clones used. Indeed, the polyclonal antibody used by Yang et al. [12] exhibited the lowest SOX10 positivity among studies featuring AdCC (88%). All cases of myoepithelioma (pooled meta-analytic estimate 75.2%, 95% CI 52.0–98.4%) were SOX10 positive when evaluated with the clone N-20 [22] and an unknown clone [21]. However, only 56% were SOX10 positive for the BC34 clone [14]. AciCC (pooled estimate of SOX10 positivity 97.4%, 95% CI 94.1–100.6%) had a crude observed SOX10 positivity rate of 96.8% and had total positivity across six studies [6, 10, 14, 22, 38, 39]. Nonetheless, only 50% were SOX10 positive in one study using a polyclonal SOX10 antibody [23]. However, Schmitt et al. [10] used the same antibody with 26 samples. With no information on the tumor grade or subtype, the reason for this difference is unknown. All samples of the intercalated subtype of intraductal carcinoma (crude observed positivity rate 100%, meta-analytic estimate 90.8%, 95% CI 75.9–105.8%) and microsecretory carcinoma (crude observed positivity rate 100%, meta-analytic estimate 96.1%, 95% CI 87.2–105.0%) were SOX10 positive. Confidence intervals exceeding 100% reflect the meta-analytic model and should not be interpreted as biologically possible results.
Strong positivity based on limited data was found in BCAC (pooled estimate 84.8%, 95% CI 67.3–102.3%) and EMC (pooled estimate 89.0%, 95% CI 71.6–106.5%), with no variations between the clones used. The crude observed SOX10 positivity rate of myoepithelial carcinoma was 100%. This finding was based on a very small sample size of two cases. There was no difference between the clones used [22, 39].
WT is essentially always SOX10 negative, with a crude observed positivity rate of 0% (pooled meta-analytic estimate 2.3%, 95% CI 0–7.0%), giving the marker a strong differential diagnostic value. The apocrine type of intraductal carcinoma (pooled estimate 10.3%, 95% CI − 10.4–31.1%), SDC (pooled estimate 3.7%, 95% CI − 0.6–8.1%), and oncocytoma (pooled estimate 8.0%, 95% CI − 2.7–18.7%) were also SOX10 negative, with a crude observed SOX10 positivity rate of 0% in all of them. In SDC, the only positive samples were low grade [6], likely representing a historical entity and thus not included in the meta-analysis. MEC can be considered mostly negative (pooled estimate 20.2%, 95% CI 2.8–37.6%), although with a crude observed positivity of 16.0% and a pooled estimate of 20.2% almost one-fifth of them are SOX10 positive. As 100% of mucoacinar subtypes and 30% of oncocytic subtypes of MEC were SOX10 positive, the subtype explains results’ variability.
Entities with considerable variability in SOX10 expression include SC (pooled estimate 79.9%, 95% CI 55.1–104.7%) and adenocarcinoma not otherwise specified (pooled estimate 39.8%, 95% CI − 4.3–84%). In SC, no subtypes or grades were reported, and the results within the SOX10 clones varied. In adenocarcinoma not otherwise specified, the only positive samples were the four studied by Dababneh et al. [21]. The SOX10 clone was not reported, and one of the entities was described as intermediate grade and the other as low grade. There was also strong heterogeneity in the SOX10 expression of intraductal carcinomas when both subtypes were combined (pooled estimate 56.4%, 95% CI 17.6–95.1%). As noted by Hsieh et al. [28], SOX10 is a strong marker of the intercalated subtype of intraductal carcinoma and is mostly negative in the apocrine subtype. This meta-analysis provides further proof of these findings and demonstrates the necessity of subtype-specific reporting in intraductal carcinomas. SOX10 expression in SP (pooled estimate 50.0%, 95% CI 0–100%) remains unclear due to insufficient data. SP was evaluated with EP268 [27] and BC34 [19], and both studies had variable results. In Hsieh et al., the oncocytic (n = 4) and exophytic (n = 5) subtypes of SP were SOX10 negative. Nine other entities were SOX10 positive. Similarly, the oncocytic SP immunostained by Owosho et al. was SOX10 negative; the other SP sample was SOX10 positive.
The differences in staining results between SOX10 antibody clones are generally negligible. However, as discussed earlier, some were found. EP268, SP267, N-20, A-2, and polyclonal antibodies produce similar results. In contrast, BC34 provides smaller percentages of SOX10-positive tumors, especially in basaloid salivary gland tumors. Regrettably, information on SOX10 staining intensity and the percentage of SOX10-positive cells was scarce. Thus, no conclusions could be drawn about the effect of the SOX10 clone on these parameters.
In the small series of cytological SOX10 stains, WT, MEC, and oncocytoma were SOX10 negative, adhering to the findings in the histological samples. However, only 62% of AciCC were SOX10 positive, highlighting a stark difference compared to the 97.4% pooled proportion of histological samples. This finding mirrors the common lowered immunopositivity in cytological preparations caused by fixation and processing differences [42–44]. The antibody used in the cytological samples was polyclonal, and the staining results are in line with the histological counterparts stained with polyclonal antibody. Of note, in histological samples, the only antibody with negative SOX10 results in AciCC was polyclonal.
To conclude the results of the meta-analysis, the list of entities likely to be SOX10 positive includes PA, BCA, BCAC, AdCC, myoepithelioma, myoepithelial carcinoma, EMC, PAC, AciCC, CA, intercalated type of intraductral carcinoma and microsecretory carcinoma. The list of entities likely to be SOX10 negative includes WT, oncocytoma, MEC, SDC, HCCC and apocrine type of intraductal carcinoma.
Although morphologic groups have overlap between each other, certain observations can be made. SOX10 tends to be positive in all entities which are likely to have basaloid morphology, limiting the usefulness of SOX10 in differentiating them from each other. In salivary gland tumors which primarily present as oncocytic, SOX10 can only be considered positive in AciCC with the rest having either high variability in results or being SOX10 negative. Salivary gland tumors with clear cell features form a heterogenous group with considerable overlap between the other groups. HCCC is mostly SOX10 negative, which helps differentiate it from myoepithelioma, myoepithelial carcinoma and AciCC. In contrast, MEC is SOX10 negative as well, limiting the usefulness of SOX10 in differentiating between HCCC and MEC.
Of the entities not further analyzed, all samples of sialoblastoma, carcinoma-ex PA, poorly differentiated carcinoma, and sclerosing polycystic adenoma were SOX10 positive. Most cases of striated duct adenoma were SOX10 positive. Only a minor portion of carcinoma NOS, SDC ex-PA, and adenocarcinoma ex-PA NOS was SOX10 positive. All samples of cribriform cystadenocarcinoma, papillary cystadenoma, lymphoepithelial carcinoma, squamous cell carcinoma, and carcinosarcoma were SOX10 negative. Due to very limited evidence, these results should be approached with caution. More research should be conducted to confirm these findings.
The most significant strength of this study lies in its large amount of data, with an analysis of 35 studies and 1183 tumors immunostained with SOX10. Twenty of the 32 histological entities were eligible for meta-analysis and forest plot construction. The weaknesses of this study stem from the use of only one database, PubMed, limiting the number of studies available for screening. Additionally, it is possible for a study to not contain the search term “SOX10” in the title, abstract or keywords, thus making it invisible in the literature search. Dozens of case reports and five studies with inconsistencies were excluded, which also limits the available data for the meta-analysis. These limitations were apparent in the original study’s design, where the number of articles proved to be insufficient for meta-analysis. Furthermore, the original study’s design would have suffered from confirmation bias due to only including the search words of entities with known SOX10 positivity. This weakness was eliminated by this study’s design, which used one search term. As non-English articles were excluded, several promising articles written in the Chinese language were excluded.
In conclusion, SOX10 is a strong and reliable marker of basaloid salivary gland tumors. However, some entities in the groups of tumors with oncocytic features or clear cell features are SOX10 immunopositive. This meta-analysis supplies a comprehensive list of SOX10-positive and -negative entities, while illuminating the need for more research on entities for which SOX10 positivity could not be discerned due to insufficient data. Furthermore, the findings highlight the necessity of subtype-specific reporting and the reporting of the antibody clone in immunohistochemical studies.
Supplementary Information
Below is the link to the electronic supplementary material.
Author Contributions
O.T. and I.K. wrote the main manuscript text. O.T., H.L., D.K. collected and analyzed data. O.T. and D.K. prepared figures and tables. I.K. conceptualized and supervised the project. All authors reviewed the manuscript.
Funding
This study was supported by state funding for university-level research awarded to Tampere University Hospital/Fimlab Laboratories, Wellbeing Services County of Pirkanmaa, and by the Cooperation Program of Charles University within the research area of Surgical Disciplines.
Data Availability
The dataset used to support the findings of this study is available from the corresponding author at ivana.kholova@tuni.fi upon reasonable request.
Code Availability
No custom code was developed or used in this study. Statistical analyses were performed using OpenMeta [Analyst] Version 0.24.1, while Microsoft Excel was used for data tabulation.
Declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Ethical Approval
For this type of study, ethical approval is not required.
Consent to Participate
For this type of study, informed consent is not required.
Consent For Publication
For this type of study, consent for publication is not required.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.WHO Classification of Tumours Editorial Board (2024) WHO classification of tumours–5th edition–head and neck tumours part A, 5th edition. IARC Press, Lyon [Google Scholar]
- 2.Iyer J, Hariharan A, Cao UMN, Mai CTT, Wang A, Khayambashi P, Nguyen BH, Safi L, Tran SD (2021) An overview on the histogenesis and morphogenesis of salivary gland neoplasms and evolving diagnostic approaches. Cancers (Basel) 13(15):3910. 10.3390/cancers13153910 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Sy AL, Hoang MP (2023) SOX10. J Clin Pathol 76(10):649–653. 10.1136/jcp-2023-208924 [DOI] [PubMed] [Google Scholar]
- 4.Koutlas IG, Dolan M, Lingen MW, Argyris PP (2019) Plasmacytoid cells in salivary pleomorphic adenoma: an alternative interpretation of their immunohistochemical characteristics highlights function and capability for epithelial–mesenchymal transition. Oral Surg Oral Med Oral Pathol Oral Radiol 128(5):515–529. 10.1016/j.oooo.2019.01.081 [DOI] [PubMed] [Google Scholar]
- 5.Dabbs DJ (2021) Diagnostic immunohistochemistry—theranostic and genomic applications. Elsevier–Health Sciences Division, Philadelphia [Google Scholar]
- 6.Hsieh MS, Lee YH, Chang YL (2016) SOX10-positive salivary gland tumors: a growing list, including mammary analogue secretory carcinoma of the salivary gland, sialoblastoma, low-grade salivary duct carcinoma, basal cell adenoma/adenocarcinoma, and a subgroup of mucoepidermoid carcinoma. Hum Pathol 56:134–142. 10.1016/j.humpath.2016.05.021 [DOI] [PubMed] [Google Scholar]
- 7.Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD et al (2021) The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 372:n71. 10.1136/bmj.n71 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Thompson LD, Bauer JL, Chiosea S, McHugh JB, Seethala RR, Miettinen M, Müller S (2015) Canalicular adenoma: a clinicopathologic and immunohistochemical analysis of 67 cases with a review of the literature. Head Neck Pathol 9(2):181–195. 10.1007/s12105-014-0560-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.North JP, McCalmont TH, Fehr A, van Zante A, Stenman G, LeBoit PE (2015) Detection of MYB alterations and other immunohistochemical markers in primary cutaneous adenoid cystic carcinoma. Am J Surg Pathol 39(10):1347–1356. 10.1097/PAS.0000000000000463 [DOI] [PubMed] [Google Scholar]
- 10.Schmitt AC, Cohen C, Siddiqui MT (2015) Expression of SOX10 in salivary gland oncocytic neoplasms: a review and a comparative analysis with other immunohistochemical markers. Acta Cytol 59(5):384–390. 10.1159/000441890 [DOI] [PubMed] [Google Scholar]
- 11.Krings G, Joseph NM, Bean GR, Solomon D, Onodera C, Talevich E, Yeh I, Grenert JP, Hosfield E, Crawford ED, Jordan RC, van Zante A, Zaloudek C, Shin SJ, Chen YY (2017) Genomic profiling of breast secretory carcinomas reveals distinct genetics from other breast cancers and similarity to mammary analog secretory carcinomas. Mod Pathol 30(8):1086–1099. 10.1038/modpathol.2017.32 [DOI] [PubMed] [Google Scholar]
- 12.Yang C, Zhang L, Sanati S (2019) SOX10 is a sensitive marker for breast and salivary gland adenoid cystic carcinoma: immunohistochemical characterization of adenoid cystic carcinomas. Breast Cancer (Auckl) 13:1178223419842185. 10.1177/1178223419842185 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Adkins BD, Geromes A, Zhang LY, Chernock R, Kimmelshue K, Lewis J Jr., Ely K (2020) SOX10 and GATA3 in adenoid cystic carcinoma and polymorphous adenocarcinoma. Head Neck Pathol 14(2):406–411. 10.1007/s12105-019-01046-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ko YCK, Varma S, Zhu CF, Zhu SX, Vennam S, Poh CF, Jordan RC, Kong C, Pollack JR, West RB (2020) Gene expression profiling of head and neck tumors identifies FOXP1 and SOX10 expression as useful for distinguishing ameloblastoma from basaloid salivary gland tumors. Am J Surg Pathol 44(5):665–672. 10.1097/PAS.0000000000001421 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Bishop JA, Sajed DP, Weinreb I, Dickson BC, Bilodeau EA, Agaimy A, Franchi A, Khurram SA, Da Forno P, Robledo J, Kalmar JR, Aguirre S, Krane JF, Tapia JL, Kiss K, Cordell K, Rosebush M, Barrett AW, Oda D, Assaad A, Nagao T, Kawakami F, Nakaguro M, Zahir I, Wakeman K, Ihrler S, Chenevert J, Lin YL, Westra WH, Gagan J, Rooper LM (2021) Microsecretory adenocarcinoma of salivary glands: an expanded series of 24 cases. Head Neck Pathol 15(4):1192–1201. 10.1007/s12105-021-01331-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Fisch AS, Laklouk I, Nakaguro M, Nosé V, Wirth LJ, Deschler DG, Faquin WC, Dias-Santagata D, Sadow PM (2021) Intraductal carcinoma of the salivary gland with NCOA4-RET: expanding the morphologic spectrum and an algorithmic diagnostic approach. Hum Pathol 114:74–89. 10.1016/j.humpath.2021.05.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.McLean-Holden AC, Rooper LM, Lubin DJ, Magliocca KR, Manucha V, Sadow PM, Tobias J, Vargo RJ, Thompson LDR, Heidarian A, Weinreb I, Wenig B, Gagan J, Hernandez-Prera JC, Bishop JA (2022) Frankly invasive carcinoma ex-intraductal carcinoma: expanding on an emerging and perplexing concept in salivary gland tumor pathology. Head Neck Pathol 16(3):657–669. 10.1007/s12105-021-01408-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Rooper LM, Agaimy A, Assaad A, Bal M, Eugene H, Gagan J, Nonogaki H, Palsgrove DN, Shah A, Stelow E, Stoehr R, Thompson LDR, Weinreb I, Bishop JA (2023) Recurrent IDH2 mutations in salivary gland striated duct adenoma define an expanded histologic spectrum distinct from canalicular adenoma. Am J Surg Pathol 47(3):333–343. 10.1097/PAS.0000000000002004 [DOI] [PubMed] [Google Scholar]
- 19.Owosho AA, Shasteen AM, Aguirre SE, Summersgill KF (2023) Clinicopathologic study of sialadenoma papilliferum of the minor salivary glands: a series of 8 new cases with BRAF V600E mutation-specific immunohistochemical analysis. Int J Surg Pathol 31(7):1265–1272. 10.1177/10668969221147170 [DOI] [PubMed] [Google Scholar]
- 20.Delfin L, Doff JJ, Gagan J, Flack A, Krane JF, Jo VY, Torell AG, Palsgrove D, Bishop JA (2024) Pure apocrine intraductal carcinoma of salivary glands: reassessment of molecular underpinnings and behavior. Head Neck Pathol 18(1):58. 10.1007/s12105-024-01653-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Dababneh MN, Azzato EM, Nakitandwe J, Cracolici V, Shah AA (2025) HMGA2-positive salivary gland neoplasms with prominent trabecular/canalicular morphology: a focus on carcinomas arising within this phenotype. Histopathology 86(3):385–396. 10.1111/his.15334 [DOI] [PubMed] [Google Scholar]
- 22.Ohtomo R, Mori T, Shibata S, Tsuta K, Maeshima AM, Akazawa C, Watabe Y, Honda K, Yamada T, Yoshimoto S, Asai M, Okano H, Kanai Y, Tsuda H (2013) SOX10 is a novel marker of acinus and intercalated duct differentiation in salivary gland tumors: a clue to the histogenesis for tumor diagnosis. Mod Pathol 26(8):1041–1050. 10.1038/modpathol.2013.54 [DOI] [PubMed] [Google Scholar]
- 23.Urano M, Nagao T, Miyabe S, Ishibashi K, Higuchi K, Kuroda M (2015) Characterization of mammary analogue secretory carcinoma of the salivary gland: discrimination from its mimics by the presence of the ETV6-NTRK3 translocation and novel surrogate markers. Hum Pathol 46(1):94–103. 10.1016/j.humpath.2014.09.012 [DOI] [PubMed] [Google Scholar]
- 24.Takamatsu M, Sato Y, Muto M, Nagano H, Ninomiya H, Sakakibara R, Baba S, Sakata S, Takeuchi K, Okumura S, Ishikawa Y (2018) Hyalinizing clear cell carcinoma of the bronchial glands: presentation of three cases and pathological comparisons with salivary gland counterparts and bronchial mucoepidermoid carcinomas. Mod Pathol 31(6):923–933. 10.1038/s41379-018-0025-7 [DOI] [PubMed] [Google Scholar]
- 25.Kobayashi K, Yoshimoto S, Omura G, Matsumoto Y, Sakai A, Eguchi K, Sakai T, Honma Y, Matsumoto F, Kawazu M, Saito Y, Ryo E, Yoshida A, Yatabe Y, Mori T (2024) An institutional experience of core needle biopsy with cooperative implementation by pathologists and head and neck oncologists in salivary gland carcinoma. Head Neck Pathol 18(1):122. 10.1007/s12105-024-01725-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Sakyo A, Ryo E, Nishino S, Kobayashi K, Yoshimoto S, Omura G, Fushimi C, Sakai T, Sakai A, Eguchi K, Takahashi H, Yokoyama K, Honma Y, Mori A, Kato H, Hatano T, Yoshida A, Matsumoto F, Yatabe Y, Mori T (2025) NKX3.1 helps distinguish hyalinizing clear cell carcinoma from other clear cell salivary gland neoplasms. Lab Invest 105(3):102205. 10.1016/j.labinv.2024.102205 [DOI] [PubMed] [Google Scholar]
- 27.Hsieh MS, Bishop JA, Wang YP, Poh CF, Cheng YL, Lee YH, Jin YT, Chang JYF (2020) Salivary sialadenoma papilliferum consists of two morphologically, immunophenotypically, and genetically distinct subtypes. Head Neck Pathol 14(2):489–496. 10.1007/s12105-019-01068-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Hsieh MS, Lee YH, Jin YT, Kuo YJ (2020) Clinicopathological study of intraductal carcinoma of the salivary gland, with emphasis on the apocrine type. Virchows Arch 477(4):581–592. 10.1007/s00428-020-02823-7 [DOI] [PubMed] [Google Scholar]
- 29.Skálová A, Agaimy A, Stanowska O, Baneckova M, Ptáková N, Ardighieri L, Nicolai P, Lombardi D, Durzynska M, Corcione L, Laco J, Koshyk O, Žalud R, Michal M, Vanecek T, Leivo I (2020) Molecular profiling of salivary oncocytic mucoepidermoid carcinomas helps to resolve differential diagnostic dilemma with low-grade oncocytic lesions. Am J Surg Pathol 44(12):1612–1622. 10.1097/PAS.0000000000001590 [DOI] [PubMed] [Google Scholar]
- 30.Skálová A, Baněčková M, Laco J, Di Palma S, Agaimy A, Ptáková N, Costes-Martineau V, Petersson BF, van den Hout MFCM, de Rezende G, Klubíčková N, Koblížek M, Koshyk O, Vaneček T, Leivo I (2022) Sclerosing polycystic adenoma of salivary glands: a novel neoplasm characterized by PI3K–AKT pathway alterations—new insights into a challenging entity. Am J Surg Pathol 46(2):268–280. 10.1097/PAS.0000000000001807 [DOI] [PubMed] [Google Scholar]
- 31.Klubíčková N, Loghides F, van den Hout MFCM, Costes-Martineau V, Ferrara G, Rito M, Hájková V, Grossmann P, Šteiner P, Kovářová I, Michal M, Leivo I, Skálová A (2025) Expanding the molecular–genetic spectrum of canalicular adenoma-like subtype of pleomorphic adenoma of salivary glands. Am J Surg Pathol 49(6):554–563. 10.1097/PAS.0000000000002377 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Agaimy A, Baněčková M, Ihrler S, Mueller SK, Franchi A, Hartmann A, Stoehr R, Skálová A (2021) ALK rearrangements characterize 2 distinct types of salivary gland carcinomas: clinicopathologic and molecular analysis of 4 cases and literature review. Am J Surg Pathol 45(9):1166–1178. 10.1097/PAS.0000000000001698 [DOI] [PubMed] [Google Scholar]
- 33.Agaimy A, Ihrler S, Baněčková M, Costés Martineau V, Mantsopoulos K, Hartmann A, Iro H, Stoehr R, Skálová A (2022) HMGA2-WIF1 rearrangements characterize a distinctive subset of salivary pleomorphic adenomas with prominent trabecular (canalicular adenoma-like) morphology. Am J Surg Pathol 46(2):190–199. 10.1097/PAS.0000000000001783 [DOI] [PubMed] [Google Scholar]
- 34.Aquino SN, Bezerra HKF, Louredo BVR, Amaral-Silva GKD, Gaetti-Jardim EC, Antunes DM, Santos-Silva AR, Lopes MA, Vargas PA (2025) Clinicopathological and immunohistochemical aspects of conventional and unicystic canalicular adenoma: a case series. Oral Surg Oral Med Oral Pathol Oral Radiol 139(3):e59–e67. 10.1016/j.oooo.2024.08.022 [DOI] [PubMed] [Google Scholar]
- 35.Weinreb I, Hahn E, Dickson BC, Rooper LM, Rupp NJ, Freiberger SN, Lubin D, Gagan J, Bishop JA (2023) Microcribriform adenocarcinoma of salivary glands: a unique tumor entity characterized by an SS18::ZBTB7A fusion. Am J Surg Pathol 47(2):194–201. 10.1097/PAS.0000000000001980 [DOI] [PubMed] [Google Scholar]
- 36.Baghai F, Yazdani F, Etebarian A, Garajei A, Skalova A (2017) Clinicopathologic and molecular characterization of mammary analogue secretory carcinoma of salivary gland origin. Pathol Res Pract 213(9):1112–1118. 10.1016/j.prp.2017.07.017 [DOI] [PubMed] [Google Scholar]
- 37.Luna-Ortiz K, Dominguez-Malagon H, Corredor-Alonso GE, Reynoso-Noveron N, Herrera-Ponzanelli C, Luna-Peteuil Z, Zacarias-Ramon LC (2022) Clinicopathological and immunohistochemical behavior of ductal carcinoma of the salivary and lacrimal gland in a Mexican Mestizo population. Eur Arch Otorhinolaryngol 279(1):327–333. 10.1007/s00405-021-06757-x [DOI] [PubMed] [Google Scholar]
- 38.Ni H, Zhang XP, Wang XT, Xia QY, Lv JH, Wang X, Shi SS, Li R, Zhou XJ, Rao Q (2016) Extended immunologic and genetic lineage of mammary analogue secretory carcinoma of salivary glands. Hum Pathol 58:97–104. 10.1016/j.humpath.2016.09.001 [DOI] [PubMed] [Google Scholar]
- 39.Lee JH, Kang HJ, Yoo CW, Park WS, Ryu JS, Jung YS, Choi SW, Park JY, Han N (2019) PLAG1, SOX10, and Myb expression in benign and malignant salivary gland neoplasms. J Pathol Transl Med 53(1):23–30. 10.4132/jptm.2018.10.12 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Bundele M, Weinreb I, Xu B, Chiosea S, Faquin W, Dias-Santagata D, Leon M, Hyrcza M, Seethala RR (2021) Mucoacinar carcinoma: a rare variant of mucoepidermoid carcinoma. Am J Surg Pathol 45(8):1028–1037. 10.1097/PAS.0000000000001752 [DOI] [PubMed] [Google Scholar]
- 41.Mahomed F, de Bruin J, Ngwenya S, Masango Z, Hodkinson K (2024) ETV6 molecular heterogeneity in salivary secretory carcinoma: a case series report and literature review. Head Neck Pathol 18(1):66. 10.1007/s12105-024-01673-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Srebotnik Kirbis I (2025) State of the art and science of immunocytochemistry. Acta Cytol 69(1):51–59. 10.1159/000539634 [DOI] [PubMed] [Google Scholar]
- 43.Lozano MD, Robledano R, Argueta A (2025) Quality assurance in immunocytochemistry: a review and practical considerations. Acta Cytol 69(1):60–68. 10.1159/000540532 [DOI] [PubMed] [Google Scholar]
- 44.Mansour MSI, Pettersson L, Pedersen R, Mager U, Hejny K, Kronqvist F, Aronsson A, Seidal T, Akyürek LM, Kholová I (2026) Choice of fixative affects programmed death-ligand 1 expression in cell blocks from pleural effusions with metastatic pulmonary adenocarcinoma. Cancer Cytopathol 134(3):e70078. 10.1002/cncy.70078 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The dataset used to support the findings of this study is available from the corresponding author at ivana.kholova@tuni.fi upon reasonable request.
No custom code was developed or used in this study. Statistical analyses were performed using OpenMeta [Analyst] Version 0.24.1, while Microsoft Excel was used for data tabulation.