FNA of the thyroid started in the late 1980s at the Philippine General Hospital, a 1,500-bed tertiary government hospital that serves as the training facility for health sciences students enrolled in the University of the Philippines. Consultants and resident physician trainees from the departments of pathology and surgery started doing FNA of various palpable lesions mainly in the neck, 40% to 50% of which were thyroid aspirates. Thereafter, an aspiration cytology unit in the Department of Pathology housed at the College of Medicine of the University was established upon the return of a faculty member of the department from a 2-month training in FNA at the Cytology Unit of Karolinska Hospital in Stockholm, Sweden, in 1987. Two years later, another pathologist with interest in cytopathology established an FNA clinic in a private university hospital in Metro Manila.

Resident physicians in the Department of Pathology used the technique prescribed by the Scandinavian pioneers utilizing a 10-mL syringe with 23G to 25G needles attached to a syringe holder. Some surgeons and endocrinologists performed the procedure without a syringe holder. Smears were prepared by the aspirationist and cytotechnicians stained the smears using either Papanicolaou stain or modified Wright-Giemsa stain (Diff-Quik), or both. All interpretations were rendered by the pathologist on duty.

In the 1990s, private hospitals in the national capital region (Metro Manila) likewise engaged in the practice of thyroid FNA with pathologists, surgeons, and endocrinologists performing the procedure. Pathologists who were trained from the university hospital performed thyroid FNA in their affiliated hospitals when they started their practice in various regions of the country.

In 2010, interventional radiologists, endocrinologists, and cytopathologists started doing thyroid FNA under ultrasound-guidance, especially for small lesions measuring from 1.0 to 1.5 cm.

Local endocrinologists follow the 2015 guidelines of the American Thyroid Association [7]. In 2011, a revisit of the clinical practice guidelines at the Philippine General Hospital for patients with thyroid cancer maintained the original 2008 recommendation of total or near total thyroidectomy for all patients with well differentiated thyroid carcinoma with nodule size of greater than 1.0 cm in diameter, and lobectomy with an isolated nodule size of less than 1.0 cm in diameter, and without lymph node metastasis on preoperative ultrasound [8]. There is no specific recommendation on preoperative FNA including terminology referable to TBSRTC. Intra-operative frozen section for non-diagnostic FNA is controversial as no consensus was reached by the multispecialty members of the panel. Likewise, international guidelines do not recommend the use of frozen section due to its limited role with high frequency of false-negative results and lack of consistent agreement between frozen section and final histological diagnosis [9].

Most practicing pathologists performing thyroid FNA use a syringe holder made of aluminum or plastic. The procedure starts with proper positioning of the patient, application of 70% alcohol at the puncture site, and localization of the lesion by immobilizing the target with one hand followed by needle puncture of the skin into the target lesion. Negative pressure or suction is applied by back and forth cutting motions within the thyroid nodule and releasing this negative pressure prior to withdrawal of the needle once the material is obtained. Some surgeons and endocrinologists use only a syringe without a syringe holder while others use the non-aspirating technique [10].

The physician-aspirationist prepares the smears by placing a small drop of the aspirated material onto a glass slide. Smear is made by laying another glass on top of the sample material and pulling the slides apart to spread it. Wet smears are fixed in 95% ethyl alcohol or air dried for submission to the laboratory. Wet smears are subjected to Papanicolaou stain while air dried smears are prepared for modified Wright-Giemsa stain (Diff-Quik). The latter stain is also used in cases where there is an on-site request for rapid evaluation of the aspirates. All smears and liquid samples obtained from cyst fluids are processed in the laboratory by cytotechnicians. Smears are routinely stained with Papanicolaou stain and the remains of the samples are processed as cell block and stained with hematoxylin and eosin. Liquid based cytology is not used for thyroid aspirates.

Formal cytopathology and cytotechnology training programs are not offered in the country at present. However, cytopathology is included as a rotation in anatomic pathology training programs. In both government and private hospitals with training in pathology, all aspirates are screened by pathology resident physician trainees and final cytodiagnosis is signed out by the pathologist on duty. In the absence of a training program, the pathologist on duty at the cytology unit screens and issues the final cytodiagnosis.

Pathologists interested in cytopathology have to study abroad for fellowship in cytopathology, after which they can apply to the local pathology society for recognition. To date, there are ten cytopathologists in the whole country, two of whom are certified by the local pathology society. Majority of the practice is in the national capital region.

There is no formal subspecialty society for cytopathology within the national organization of pathologists. Lectures in cytopathology are incorporated in the scientific program during the annual convention of the Philippine Society of Pathologists. Within individual training institutions, conferences in cytopathology are embedded in the training program of anatomic pathology.

Prior to the 2009 publication of TBSRTC [11], which is a standardized six-tier nomenclature, the aspiration biopsy cytology method was used [4,5]. This method provides a spectrum of diagnosis that closely resembles the histopathologic diagnosis of thyroid disorders. In most instances, it is cytologically possible to categorize a thyroid lesion into one of the three main entities: hyperplasia/adenomatous colloid nodule, thyroiditis, and neoplasm. There are also indeterminate cases in which the presence of cellular atypia cannot totally rule out a malignancy. For this, “atypia suspicious for malignancy” cytodiagnosis is given and followed by a recommendation for clinical correlation or for further investigation if warranted. Cytodiagnosis is based on the assessment of cellularity, architectural or group patterns of diagnostic cells, individual cell cytology including nuclear and cytoplasmic characteristics, and the presence of other cells and materials in the background. Criteria for non-diagnostic aspirates (category I) follow the criteria prescribed by TBSRTC, and these include bloody samples, presence of blood and colloid only, paucicellular smears with less than six follicular cell groups of ten cells each, and poorly prepared smears [11]. Aspiration performed with dissolution of the nodule is compatible with a thyroid cyst and is not considered as non-diagnostic if pathologist has a clear knowledge of the patient’s biopsy findings. With the introduction of TBSRTC in 2009, a majority of the thyroid FNA reports utilized both descriptive diagnosis and its corresponding equivalent category number in the six-tier nomenclature for a more effective understanding by the referring physicians.

Difficult thyroid aspirates are handled in various ways by pathologists. Self-review is done with subsequent intradepartmental referral to another pathologist if a cytopathologist is not part of the hospital staff. Presently, the Philippines is still in the process of developing an external quality assurance program in cytopathology.

Based on a recent survey of FNA practice in 16 hospitals in the country, thyroid FNA comprised 46% to 85% of the total FNA in five major hospitals at the national capital region, and 30% to 85% in 11 hospitals in various regions of the country. At the Philippine General Hospital, 60% of aspirates were from the thyroid. Being a training hospital, it has the highest number of non-diagnostic aspirates, ranging from 37% to 46% with an average of 42%. The high number is unique since different resident physician trainees perform the procedure in the course of their rotation in the departments engaged in FNA. Four private hospitals from the national capital region with a bed capacity from 300 to almost 1,000 had a non-diagnostic yield ranging from 5% to 7%. In the various regions of the country from north to south, category I registered 0 to 19% from a survey of four governments and seven private hospitals with a 300–400 bed capacity in the former and 90–680 bed capacity in the latter. Category I was 0% in one private hospital in southern Philippines since the physician-aspirationist is a pathologist who always makes a rapid on-site evaluation as part of FNA. Benign aspirates (category II) comprised 50% to 85%. Aspirates in the category III accounted for 1% to 20% and were re-aspirated following the recommendation of TBSRTC. Category IV aspirates ranged from 1% to 14%, category V from 1% to 10%, and category VI from 2% to 20%.

Patients nowadays are highly mobile and may seek medical services in different institutions. It often happens that FNA is done in one hospital and thyroid surgery with histopathologic diagnosis is performed in another center. However, all thyroid surgeries with previous FNA done in the same hospital are reviewed as part of the hospitals’ monitoring program. No data has been released but may soon be forthcoming.

Ancillary studies of thyroid FNA are infrequently used. Immunohistochemistry in the diagnosis of thyroid tumors such as galectin-3, cytokeratin 19, and HBME-1 are only rarely utilized in a few hospitals at the national capital region for thyroid surgical pathology specimens and not in aspirates. Molecular testing for somatic mutations is yet to be developed because most of the patients opt thyroid surgery with just regular follow-up [12,13]. Health care insurance does not cover molecular testing for somatic mutations. If requested, molecular testing is mainly physician driven. If the patient agrees to pay for the molecular testing, the test is referred to neighboring Asian countries or to North America.

Researches on thyroid cytology in the country are mainly cytohistologic correlations with sensitivity, specificity, and accuracy rates. Cytohistologic correlation from three hospitals in the national capital region yielded a wide range of sensitivity from 30.7% to 73%, specificity from 83% to 100%, and accuracy from 72.8% to 87.2% [14-16]. The low sensitivity can be attributed to poor tissue sampling since most thyroid FNAs are done by palpation only. A majority (85%) of the discordant cases were due to sampling errors attributable to dual pathology with a dominant benign lesion, missing a small malignant focus which is usually a papillary microcarcinoma [14]. The reliability can be improved if FNA is guided by ultrasound as attested in both international and local studies [16-18]. Solid nodules with microcalcifications and irregular margins which were found to be significant predictors of malignancy can be readily detected [19,20]. Improvement in diagnostic yield and accuracy was observed from 82.6% to 86.2% with ultrasound guidance [16]. When combined with frozen section examination, ultrasound-guided FNA for concordant cases approached an accuracy of 97.2% in one study [21].

The incidence of each diagnostic category of the Bethesda system is rarely addressed in local studies (Table 1). Summary of available publications on the malignancy risk in the Bethesda categories is presented in Table 2. An overall risk of malignancy for malignant and suspicious for malignant cytodiagnosis is 71.4% while the overall risk of malignancy for benign cytodiagnosis is 26.5% [14]. The risk of malignancy for category III ranges from 35.3% to 50% [15,22,23]. which is higher than TBSRTC reference rate. Conveying these rates to our clinical colleagues with the recommendation to do a repeat FNA should be reconsidered for optimal patient care [22,23].

Although thyroid FNA was introduced more than 30 years ago, a centralized database may be difficult to construct because the FNA practice in the Philippines is highly individualistic. If teaching and training of FNA are structured into the major training programs of physicians, a standardized FNA protocol, particularly in specimen sampling, can be improved. Looking forward, there is a need to establish a national registry from various hospitals and training institutions. Quality assurance and other related issues can be addressed through the establishment of a local society of cytopathologists in order to develop and improve the art and science of FNA. Finally, collaborative networking amongst pathologists, endocrinologists and surgeons is a strong impetus toward unified practice in the diagnosis of thyroid disorders using FNA.