Fine-Needle Aspiration of the Salivary Glands

Major salivary glands are easily accessible; therefore, they are optimal targets for fine-needle aspiration (FNA). This technique has been used for decades in Europe, but it is still controversial in the United States. The large numbers of benign and malignant tumors that can originate in the glands, as well as the lack of cytological markers to accurately diagnose the various tumors, often foster uncertainty in FNA diagnosis of salivary gland masses. Therefore, the primary challenge of FNA is differentiating benign from malignant disease and then differentiating the various malignancies.

In 2015, an international group of pathologists developed an evidence-based tiered classification system for reporting salivary gland FNA specimens called the Milan System for Reporting Salivary Gland Cytopathology (MSRSGC). The MSRSGC consists of 6 diagnostic categories, each of which is associated with a risk of malignancy and management recommendations: non-diagnostic, non-neoplastic, atypia of undetermined significance (AUS), neoplasm (benign or salivary gland neoplasm of uncertain malignant potential [SUMP]), suspicious for malignancy (SM), and malignant. The goal of the MSRSGC is to improve communication between cytopathologists and treating clinicians as well as facilitate cytologic-histologic correlation, sharing of data from different laboratories for quality control, and research. ^[1]  

Although several papers indicate that the sensitivity and specificity of fine-needle aspiration (FNA) and frozen section of salivary glands are similar, malignant salivary gland lesions have always been difficult to classify with FNA alone.

Frozen section diagnosis should be more effective than FNA in diagnosing salivary gland tumors; unfortunately, it has a less-than-perfect track record. In a 1987 report, Conley notes that pathologists change their frozen section diagnosis in 10% of cases. ^[2] In a study of 256 salivary tumors, errors in judging malignant versus benign tumors with frozen section diagnosis occurred in 5% of cases. In 4 cases (1.6%), a malignant frozen section diagnosis was changed to benign with permanent sections. This is the most important error because it could result in more radical resection than necessary. The authors reiterated the well-known notion that radical surgery is imprudent based on the frozen section findings alone.

In a similar study of 75 patients (14 malignant), no benign lesions were overdiagnosed as malignant by frozen section examination. In 4 cases, a benign diagnosis on frozen section analysis was revised to malignant with permanent section examination. Although overall diagnostic accuracy was only 85%, no patient underwent unnecessary surgery, and 4 patients benefited from more radical surgery based on the frozen section diagnosis. This study shows that exceptions can be made to the rule that radical salivary gland surgery should never be performed on the basis of frozen pathology alone.

Although frozen section has limitations, it makes the 2-dimensional architecture of the lesion available for examination. FNA would be expected to offer an even lower accuracy because this advantage is not possible in cytologic examination. In large academic centers, FNA can correctly diagnose the cell type in 61-94% ^[3] of cases and can correctly differentiate between malignant and benign tumors in 81-98% of cases. ^[4]

A study of 220 cases of parotid gland fine-needle aspiration (FNA) with histological follow-up was conducted to compare relative accuracies of both FNA and frozen section. Frozen section was performed in 57 of these cases. Sensitivity, specificity, and accuracy for FNA when diagnostic were 86%, 92%, and 90%, respectively. Frozen section was able to change 4 malignant diagnoses by FNA to benign and provide diagnosis for 5 of the 12 nondiagnostic FNAs. The sensitivity, specificity, and accuracy of frozen section were 77%, 100%, and 88%.

The authors conclude that FNA and frozen section have similar accuracies. FNA is more sensitive, and frozen section is more specific; therefore, both techniques are complementary to each other in the diagnosis of salivary gland lesions. ^[4]

Fine-needle aspiration (FNA) enables high accuracy in identifying the nature of a lesion ^[5] and can provide the following information:

• Differentiation of inflammatory from neoplastic disease: This is especially important in patients who are immunosuppressed.

• Culture for suspected infectious masses

• Differentiation of benign from malignant disease

• Differentiation of the specific tumor cell type

• Determination of site of origin, ie, whether the tumor has arisen from the parotid or if it is metastatic

• Squamous cell carcinoma evaluation: Squamous cell carcinoma can be accurately diagnosed, and treatment can be planned based on FNA findings. A highly cellular mucoepidermoid carcinoma may appear to be squamous cell carcinoma by FNA cytology. This difference is purely academic and does not change the treatment.

• Malignancy determination: The pathologist is usually able to very accurately render an opinion about the malignant or benign nature of the lesion. This is a great help in promptly guiding treatment, and this is further addressed below.

The use of FNA allows for immediate assessment of the lesion in an outpatient setting.

• In the author's institution, most FNAs of the salivary glands are performed in an outpatient setting.

• Immediate assessment of the material is usually possible within 15 minutes.

• This initial assessment is helpful in relieving the anxiety of the patient and aids in clinical decision-making.

• If the initial aspiration is not satisfactory or other studies are needed (eg, flow cytometry, lymphoma), additional aspirations are often necessary. ^[6]

Clinical implications

The use of FNA in working up salivary gland tumors is also cost effective. Two recent studies have shown that the use of FNA in salivary gland tumors is cost-effective and diagnostic. ^{[7, 8]} Specifically, Layfield et al demonstrated that the routine use of FNA in the work-up of salivary gland lesions saves up to $70,000 per 100 patients. At the same time, fine-needle aspiration reduces the operative intervention by 65% in submandibular masses and by 35% in parotid masses.

Depending on the clinical situation, the cytopathologist, surgeon, or radiologist may perform the fine-needle aspiration (FNA). Identification of the nodule is essential to the successful aspiration. Deep lesions that are less than 1 cm in diameter are best aspirated under ultrasonographic or CT scanning guidance. ^[9]

The skin is cleaned with an alcohol swab. A 25- or 23-gauge needle is used on a plastic disposable 10- or 20-mL syringe attached to a plastic or metal holder. The nodule is immobilized between the fingers, and the needle tip is rapidly directed through the skin into the nodule. Once the needle enters the mass, the needle is continuously aspirated while the needle is rapidly moved in and out to obtain the sample. Suction is then relieved, and the needle is withdrawn and detached from the syringe. Air is aspirated, and the material is expelled on glass slides (see the images below). The material is gently but rapidly smeared on the slides and immediately dipped in alcohol fixative. Some of the slides can be left to air dry for Diff-Quick staining and rapid evaluation. The fixed material is stained by Papanicolaou or hematoxylin-eosin stains, which are both excellent for nuclear detail.

If necessary, a cellblock is prepared from cells entrapped in a blood clot or tiny tissue particles obtained by FNA. A cellblock enables the pathologist to examine the tissue similarly to a biopsy, and multiple sections can be obtained from paraffin-embedded material for special studies (immunohistochemistry). The syringe is rinsed into a container with formalin for adequate fixation of cellblock material. ^[10]

Fine-needle aspirations (FNAs) are safely performed worldwide. The complications of salivary gland FNAs are rare, but 6 potential complications exist, including the following:

• Needle track contamination by malignant cells: Numerous studies indicate that needle tract contamination by malignant cells is a rare complication despite thousands of FNAs performed worldwide. A positive correlation exists with number of passes and needle size.

• Local hemorrhage: Because major salivary glands are in close proximity to the great vessels of the head and neck, local hemorrhage due to piercing of these vessels is possible but very unlikely. Physicians can prevent hematoma formation by applying firm pressure in the area of aspiration immediately after the procedure.

• Infection: This risk is no greater than that of venipuncture and is closely correlated with the patient's immune status. Adherence to sterile techniques and cleaning the skin with alcohol minimizes this risk.

• Warthin tumor (papillary cystadenoma lymphomatosum or adenolymphoma) has a high predisposition for parotitis from FNA due to a combination of cystic spaces surrounded by oncocytic cells and poor blood supply. ^[11]

• Syncope: Some patients are prone to vasovagal reactions. Physicians should be prepared for this complication. Perform aspiration while the patient is lying down or sitting.

• Dissemination of the dislodged tumor cells through lymphatics and blood vessels: This risk is certainly lower in FNA than in incisional biopsy. ^[10]

The general contraindications to fine-needle aspiration (FNA) include the following:

• Bleeding diathesis: Because control of local hemorrhage depends on hemostasis and clotting, bleeding diathesis is a contraindication.

• Uncooperative patient: Most nondiagnostic FNA findings result from noncooperation by the patient. Prior to FNA, the physician should calm the apprehensive patient.

• Skin infection in the area of FNA ^[10]

Fine-needle aspiration accuracy

Before making a definitive treatment decision, combining fine-needle aspiration (FNA) results with other data is prudent clinical practice. A typical example is a cytologic diagnosis of a lymphoepithelial cyst. The cyst is common in patients who have immunodeficiency virus (HIV); these patients tend to have bilateral lymphoepithelial cysts, and their findings are typical on CT scanning. Typically, these patients can be managed by serial examinations and occasional serial therapeutic aspiration. Lymphomas may also be accurately diagnosed if adequate tissue can be obtained for flow cytometry. A patient with an FNA diagnosis of Warthin tumor may be best served by having a technetium scan confirmation without surgery if significant medical contraindications to surgery exist.

Recurrent tumors may also be accurately diagnosed using FNA. ^[12] Some situations exist where a benign FNA cytologic appearance should be viewed with suspicion. If the mass is painful, appears to be invading surrounding structures, or co-exists with an ipsilateral facial nerve palsy and the FNA is benign, then the mass may be malignant despite a benign cytologic appearance.

Although FNA is very accurate, situations exist where the final permanent pathology differs from the FNA result. The following statistics demonstrate the high level of sensitivity and specificity of FNA in salivary gland lesions.

In 6249 cases of primary salivary gland lesions, FNA biopsies revealed 4642 benign lesions and 1607 malignant lesions. The overall diagnostic accuracy of FNA was 48%. Sensitivity for detecting malignant and benign tumors was 73% and 91%, respectively. Results of FNA biopsy are as follows (Interlaboratory Comparison Program in Nongynecologic Cytopathology [NCG] data from1999-2003):

Table. Interlaboratory Comparison Program in Nongynecologic Cytopathology [NCG] data from 1999-2003 (Open Table in a new window)

Benign lesions that were most often misdiagnosed as malignant were monomorphic adenoma (53% false positive rate), intraparotid lymph node (36%), oncocytoma (18%), and granulomatous sialadenitis (10%). The malignant lesions that were most often misdiagnosed as benign were lymphoma (57%), acinic cell carcinoma (49%), low-grade mucoepidermoid carcinoma (43%), and adenoid cystic carcinoma (33%). ^[13]

Because of the referral pattern at these institutions, the distribution of the salivary gland lesions summarized here does not necessarily reflect the true incidence in the general population.

Pitfalls of diagnosis in salivary gland fine-needle aspiration

See the list below:

• Sampling error: Correct identification and immobilization of the lesion should allow the pathologist to avoid undersampling the lesion. Having the pathologist who performs the FNA also interpret the lesions is a tremendous advantage.

• Interpretation error: Pathologist experience minimizes this error.

• Bias: Pathologist bias, augmented by the clinician's opinion, has been shown to lead to errors in diagnosis. ^[14]

• Lack of flow cytometry use: Many of the false-positive and false-negative interpretations of lymphoma could have been prevented with flow cytometry.

• Lack of immunophenotyping studies: These studies can improve interpretation of low-grade lymphoproliferative processes.

• Technical problems: Air-drying is one of the most common technical problems. Romanowsky-type stains are superior to Papanicolaou stain for assessing stromal elements on salivary gland FNAs. ^[13]

Role of special techniques

Some tumors have unique characteristics that can be discovered using special techniques. Although these are helpful, they are not always completely diagnostic.

• Special stains

  • Mucin stains (eg, mucicarmine, periodic acid-Schiff [PAS]) can identify mucin in mucoepidermoid carcinomas.

  • PAS can identify the glycogen in acinic cell carcinomas.

• Immunohistochemistry

  • S100: Some spindle cell myoepitheliomas may be very difficult to diagnose without help of positive S100 immunostains.

  • Cytokeratin: Cytokeratin confirms the epithelial nature (ie, versus melanoma, lymphoma, sarcoma) of a tumor.

  • Vimentin: Vimentin confirms the immuno-viability of cells.

  • Leukocyte common antigen (LCA): LCA confirms malignant lymphoma.

• Electron microscopy (EM): EM can identify mitochondria in myoepitheliomas.

• Flow cytometry: Flow cytometry is particularly useful in primary diagnosis and follow-up of hematolymphoid neoplasms.

Fine-needle aspiration (FNA) of the salivary glands is a safe and reliable technique in the primary diagnosis and follow-up of patients. This technique is used with great success in Europe and is gaining increased popularity in the United States as the experience of pathologists and confidence of clinicians increase. With the application of special techniques, FNA can be a very important part of clinical decision-making and preoperative counseling in most cases.

As always, the patient's overall status must be assessed, and an individual treatment plan must be created in each case. Although patients with high cardiac risk may be best served by observation given a benign FNA diagnosis, young healthy patients may require surgical intervention regardless of a negative FNA finding. A negative FNA finding alone should not prevent surgical intervention when it is otherwise clinically indicated. Similarly, FNA alone cannot dictate the extent of surgery because that judgment is based on tumor location and patient status.

Sacrifice of the facial nerve or other structures is generally best guided by findings at surgery rather than mindless obedience to preoperative dictums, as noted by Spiro: "Facial nerve preservation is ill advised when a parotid tumor has to be transected to spare the nerve. This rule applies even when the tumor is benign." This is not to say that decisions are always easy. On the contrary, making on-the-spot decisions with unusual tumor types that cannot be diagnosed with certainty at the time of surgery is often difficult.

Ultimately, diagnosis and treatment depend upon the experience of the pathologist and surgeon when dealing with anything less than permanent section diagnosis. In a center that has a relatively large volume and a pathologist with a great deal of experience and interest in salivary gland tumors, FNA and/or frozen section diagnosis may be adequate to make preoperative and operative decisions.