Faculty Bibliography

Background: The American College of Radiology (ACR) 2017 Thyroid Imaging Reporting and Data System (TI-RADS) added a new risk stratification system for classifying thyroid nodules based on sonographic appearance (T1-T5). FNA is generally not recommended for benign or low suspicion nodules. However, other factors such as nodule size and family history may trigger an order for an FNA. Our study aimed to examine the cytologic diagnosis, molecular profiles and surgical follow up in a select group of patients with sonographically benign appearing thyroid nodules.
Design(s): We performed a retrospective review in our pathology database of cases from 1/1/2016-4/1/2018, prior to our institution's adoption of the TI-RADS classification system. Thyroid nodules with in-house ultrasound exam (US), FNA cytology, The Bethesda System (TBS) cytology diagnosis, molecular testing, and surgery were included. The USs from these cases were retrospectively reviewed and assigned TI-RADS scores (TR1-TR5) by a board certified radiologist. There were no TR1 nodules. TR2 (not suspicious) and TR3 (mildly suspicious) nodules were selected for evaluation.
Result(s): From 1/1/2016-4/1/2018, there were a total of 34 patients that fit the selection criteria. Of these, there were 5 TR2 thyroid nodules and 29 TR3 thyroid nodules with corresponding FNA TBS, molecular and surgical diagnoses (table1). (Table presented)
Conclusion(s): Our study shows that sonographically benign appearing/low suspicion thyroid nodules may display molecular alterations; 50% of those proved to be RAS mutations in our study. Approximately 60% of aspirated TR2 nodules and 66% of TR3 nodules were malignant or NIFTP on excision. Despite their lower suspicion index on US, with lower TI-RADS scores, benign appearing nodules on US need to be evaluated in the context of clinical, cytologic and molecular information in order to determine clinical course

Background: Salivary gland neoplasms are rare and the majority are benign with only 20% displaying malignancy. Fine needle aspiration (FNA) plays an essential role in the initial evaluation of salivary gland lesions by providing a pre-operative diagnosis to determine appropriate management. Recently, a tiered classification system known as the Milan System for reporting Salivary Gland cytopathology (MSRSGC) has been published. This system formalizes diagnostic categories with related malignancy risk, recommended clinical therapy and follow-up. Our study aims to compare sensitivity, specificity and risk of malignancy (ROM) between the MSRSGC and the original FNA cytology diagnostic categories used at our institution to determine if the MSRSGC offers added benefit.
Design(s): Salivary gland cytology slides from subjects with final surgical pathology resections from 11/2016-06/2019 were blindly reviewed and classified according to the MSRSGC. MSRSGC diagnoses were correlated with surgical pathology diagnoses and compared to the original cytology diagnostic categories. Sensitivity, specificity and ROM of diagnostic categories were calculated for both systems.
Result(s): Follow-up histopathology was available for 101 patients with salivary gland lesions. The MSRSGC had a sensitivity of 69.0% and a specificity of 92.9%. The original classification system had a sensitivity of 75.0% and a specificity of 89.9%. ROM for MSRSGC categories and original diagnostic categories are given in Table 1 and listed side by side to reflect distribution of cases in each system. (Table presented)
Conclusion(s): Performance of the MSRSGC was comparable to that of the original classification system in the majority of cases. Both systems had a similar sensitivity, specificity and ROM in the equivalent categories. The single "non-diagnostic" and the three "nonneoplastic" cases under MSRSGC that showed histopathologic evidence of malignancy were called "negative for malignancy" in the original classification showing lack of cytohistologic correlation for both systems due to sampling errors. Two of the three cases classified as "atypia of undetermined significance" under the MSRSGC were originally classified as "negative for malignancy". Our findings suggest that traditional diagnostic classification methods for salivary gland cytopathology already established at an institution can perform as well as the MSRSGC in relaying the appropriate diagnostic information, undermining the need for transition to a new classification system

Background: HPV-related oropharyngeal squamous cell carcinoma (OP-SCC) has a superior prognosis and response to therapy than that of conventional head-and neck SCC (HNSCC). The College of American Pathologists (CAP) guidelines recommend that P16 immunostaining (IHC) in >70% of tumor cells is an excellent surrogate marker for HPV in surgical pathology OP-SCC. Fine needle aspiration (FNA) cytology is an ideal method for obtaining diagnostic material for OP-SCC and may represent the only attainable specimen. However, there is no consensus for interpretation of P16 IHC result in cytology preparations. Our study aims to assess OP-SCC P16 staining in cell block cytology preparations in comparison with P16 staining on surgical pathology specimens.
Design(s): FNA specimens from 2014-2019 of OP-SCC with P16 IHC staining were obtained. Surgical pathology P16 IHC results were set as the gold standard. Cytology cell block tumor cellularity (<100 vs >100 cells) and P16 percentage of tumor cell positivity (0%, 1-10%, 11- 50%, 51-70%, and >70%) were recorded. Using different threshold levels of P16 tumor cell positivity in cell blocks as compared with surgical P16 IHC results, sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated.
Result(s): 40 matched FNA neck lymph node/mass cytology and surgical cases were identified. Sensitivities and specificities varied when thresholds changed, with sensitivities and specificities ranging from 93.5% and 66.7% (respectively) when any P16 positivity is seen (>0%), to 56.7% and 100% (respectively) when P16 positive threshold is set at >70% (table 1 and figure 1). <100 and >100 tumor cells were seen in 11 and 29 cases respectively. (Table presented)
Conclusion(s): Our study shows that P16 IHC performed on cytology cell blocks can serve as a surrogate marker for the detection of HPV, similar to P16 staining in surgical pathology, with high sensitivity and specificity levels. The challenge in cytology specimens is choosing the proper threshold to balance between the optimal sensitivity and specificity. Our data suggests that using a threshold lower than that of surgical pathology (70%) for p16 positivity may be appropriate for FNA specimens, as lower thresholds displayed increased sensitivities with only moderately lower specificities. Of note out of the 11 cases with <100 tumor cells, only one cases was a false negative, indicating that tumor cellularity may not affect P16 interpretation on cell block

Background: TPS is a reporting system that includes specific diagnostic categories and cytologic criteria for the accurate diagnosis of high-grade urothelial carcinoma (HGUC). Its success is dependent on its acceptance and widespread use by the cytology and urology communities. Since its development in 2013, institutions have been transitioning to TPS in an effort to standardize terminology and increase the sensitivity of diagnosing HGUC. We present our data comparing TPS diagnoses (PD) to the traditional reporting system (TD) in correlation with the gold standard surgical pathology diagnosis (SD).
Design(s): A search of the pathology database was conducted on urine cytology specimens from adult patients from 7/1/2014-6/30/2016 (TD) and 7/1/2017-6/30/2019 (PD). 454 cytology specimens from 382 patients were found to have corresponding urinary tract SD within 90 days and were included in the study. 192/454 urines were from prior to TPS implementation; 262/454 were from after TPS implementation. TD included: Positive for malignancy/urothelial carcinoma (POS), suspicious for malignancy/urothelial carcinoma (SUS), atypical, low-grade urothelial neoplasia/carcinoma (LG), and negative for malignancy (NEG). TD and PD were compared to their corresponding SD.
Result(s): 34/41 (83%) of HGUC were correctly identified using PD compared to 36/49 (73%) using TD. 15/23 (65%) of SHGUC correlated with HGUC on SD using PD compared with 13/16 (81%) with TD. Rates of "Atypical" diagnoses were decreased from 82/192 (42%) with TD to 95/262 (36%) with PD while the risk of malignancy (ROM) with "Atypical" diagnosis increased using PD from 33% to 36%. LG was identified on cytology in 1/43 (2%) using TD and 3/65 (5%) with PD. In both TD and PD, LG cytologic diagnosis had 100% specificity. 31/65 (48%) of LGUN were correctly categorized as NHGUC using PD while 10/43 (23%) were NEG in TD. (Table 1 and Figure 1) (Table presented)
Conclusion(s): Implementation of TPS in our laboratory led to a higher accuracy in the cytologic diagnosis of HGUC. Additionally, the "Atypical" rate decreased from 42% to 36% while the ROM showed a modest increase. While 12% of HGUCs diagnosed with TPS were found to be benign on SD, 60% of these cases actually had prior and/or subsequent HGUC/CIS on SD indicating that original PD was in fact concordant. LGUN is difficult to diagnose on cytology, and TPS afforded an increase in NHGUC diagnoses in line with the main goal of the PD- diagnosis of HGUC

BACKGROUND:Differentiating parathyroid from thyroid lesions can be difficult on fine-needle aspiration (FNA) due to overlapping cytomorphologic features. While the traditional parathyroid hormone (PTH) assays can help in the distinction, these tests may be cumbersome, particularly when the lesion is unexpected clinically and a needle wash is not collected at the time of FNA. Therefore, we chose to investigate the application of immunohistochemical staining (IHC) with GATA 3 and thyroid transcription factor-1 (TTF-1) on air-dried cytology smears to distinguish parathyroid and thyroid lesions. METHODS:Air-dried touch preparation (TP) slides were prepared from consecutively selected parathyroid and thyroid specimens. Thirteen FNA cases with the clinical concern for parathyroid lesions were also included in the study. IHC was performed on unstained and ultrafast Papanicolaou (UFP) stained air-dried slides. RESULTS:On TP slides, GATA 3 expression was observed in all cases of parathyroid origin but no immunoreactivity was present in thyroid lesions. TTF-1 expression was observed in all cases of thyroid origin but not in parathyroid lesions. GATA 3 and TTF-1 expression of 13 FNA cases were consistent with the clinical impression or concurrent PTH tests. CONCLUSIONS:IHC with GATA 3 and TTF-1 on air-dried cytology smears is a simple and effective way to differentiate parathyroid vs thyroid lesions on FNA. Air-dried unstained and UFP-stained slides perform equally well with IHC, but UFP-stained slides provide the added benefit of morphologic evaluation and assessment of smear cellularity prior to IHC.

BACKGROUND:Hurthle cell lesions often pose diagnostic challenges, despite their common occurrence on thyroid fine-needle aspiration cytology (FNAC). The associated molecular alterations are also not well understood. Therefore, our study aimed to delineate the molecular profile of Hürthle cell lesions classified as Bethesda Categories III or IV (atypia of undetermined significance (AUS) or suspicious for follicular neoplasm (SFN)) on FNAC and to correlate this molecular profile with surgical resection findings. METHODS:This study consisted of 188 Hürthle cell lesions with indeterminate cytology and ThyroSeq® v2/v3 molecular testing results. Surgical follow-up was available for 33 cases. RESULTS:The majority of indeterminate Hürthle cell lesions had negative ThyroSeq® results (61%) and were benign on available surgical follow-up. The most prevalent mutations involved the RAS gene (21%), which were associated with benign lesions, non-invasive follicular thyroid neoplasms with papillary-like nuclear features (NIFTP), and malignancy. The remaining mutations involved less than 18% of the cases, including PAX8/PPARG (3.7%), TSHR (3.7%), EIF1AX (2.7%), MET (2.1%), PTEN (1.6%), clonal copy number alteration (1.6%), TERT (1.1%), and 0.5% each of GNAS, PIK3CA, and TP53 mutations. On follow-up, 45% were benign, 24% were NIFTP, and 30% were malignant. The malignant cases had different molecular alterations. CONCLUSION/CONCLUSIONS:No single molecular alteration defines cytologically indeterminate Hürthle cell lesions; the majority of cases have low-risk or no molecular alterations and are benign on follow-up. These findings suggest that molecular testing may be useful, but is not definitive, in determining which cases may be managed conservatively; additional studies are needed to fully determine the negative predictive value in ruling out malignancy.

Introduction: TBS diagnostic category rates, ASCUS/(+)hrHPV (high risk HPV) ratio, and cytotechnologist's (CT'S) concordance with the CP's final diagnosis are used as common quality monitors in gyn cytology. Additionally, extending monitoring of the hrHPV (+) rate to NILM and SIL cases has been proposed as quality indicators for cytopathologist's (CP's) performance. At our institution, Pap tests are finalized without the knowledge of hrHPV status. We investigated the impact of known hrHPV status on CPs' interpretation of cases previously screened as NILM, and stipulated its potential consequence on quality metrics. Material(s) and Method(s): 60 Pap tests previously resulted as NILM, half hrHPV (+) and half hrHPV (-), were reviewed blindly by 5 CPs in two rounds at 4 months interval. At first round, correct hrHPV results were provided to the CPs. At second round, incorrect (reversed) hrHPV results were given. McNemar chi-squared test was used to analyze the impact of knowing the hrHPV test result on Pap test interpretation. Kappa coefficient was calculated to test intra-observer agreement between the first and second review of the same slides for each CP. Result(s): ASCUS (13%) was the most upgraded diagnosis followed by 12 LSILs (2%) and 2 HSILs (0.3%). There were no significant differences in Pap test interpretation based on hrHPV status for 3 CPs and marked differences for 2 CPs (Table 1). Intra-observer agreement between round 1 and round 2 diagnoses varied from moderate to poor (Table 2). Conclusion(s): Knowledge of hrHPV status significantly biases some but not all CPs. hrHPV (+) to NILM, ASCUS and SIL ratios may not be the most objective parameters for evaluation of CP performance under these circumstances. This bias has further implications for CT performance evaluation because it impacts CT discordance rate measured against CPs final diagnosis. [Figure presented] [Figure presented]

Introduction: The Thyroid Imaging Reporting and Data System (TI-RADS) was designed to standardize risk stratification of thyroid nodules by ultrasonographic criteria and categorize nodules as TR1-TR5 to designate nodules for fine needle aspiration (FNA) or surveillance. Thyroid FNAs are classified according to The Bethesda System for Reporting Thyroid Cytopathology (TBSRTC) (categories TBS I-VI) with an associated risk of malignancy and management guideline. We utilize Thyroseq-V3 molecular testing for indeterminate cytology cases (TBS III- V). Our aim was to correlate indeterminate thyroid FNAs with TI-RADS scores and molecular results to determine if TI-RADS is accurately identifying nodules for biopsy.
Material(s) and Method(s): A retrospective review of thyroid nodules from 1/1/2018-8/30/2018 was performed. Patients with ultrasound (US) reports including TI-RADS scores, FNA reports with indeterminate cytology (TBS-III, TBS-IV and TBS-V) and molecular testing were included.
Result(s): 370 of 1000 thyroid nodules had US reports with TI-RADS scoring and concurrent cytology. 47 cases had indeterminate cytology (TBS-III n=37, TBS-IV n=7 and TBS-V n=3) and reflex molecular testing. Majority were TR4 (31/47;65.97%) and TR5 (10/47;21.27%) (Table1). 23/47 (48.94%) showed no alteration. NRAS was the most common alteration (8 cases), followed by Copy Number Alterations (CNA) (6 cases) (Figure 1). Three TBS-III cases showed dual alterations (NRAS/CNA x2 and HRAS/CNA). Two TBS-IV cases had multiple alterations (EIF1AX/NRAS/TP53 and NRAS/PTEN).
Conclusion(s): While majority of thyroid nodules had a high TI-RADS score (TR4 or TR5), most cases fell into the atypical category (TBS III). Almost half of the thyroid nodules lacked any molecular alterations thereby suggesting an over-classification by TI-RADS. Further refinement of the TI-RADS criteria may be warranted. [Figure presented] [Figure presented]
Copyright

OBJECTIVE:Because of the indolent nature of Noninvasive Follicular Thyroid Neoplasm with Papillary-Like Nuclear Features (NIFTP) and potential requisite for conservative treatment, it is crucial to identify features of this entity pre-operatively. Our group recently published our findings that there are several cytomorphologic features that may be used as clues to distinguish NIFTP, PTC and follicular adenoma (FA) on fine-needle aspiration (FNA). Therefore, we aimed to determine the interobserver reproducibility of these findings. METHODS:Pre-surgical FNA slides from NIFTP (n=30), classic PTC (n=30) and FA (n=30) collected from 1/2013-8/2016 were reviewed by 7 cytopathologists blindly. Presence of selected cytomorphologic features was recorded and compared to determine percent agreement and inter-rater reliability among study cytopathologists using Gwet's AC1 statistics. RESULTS:For all the cytomorphologic features, the overall percent agreement amongst the pathologists ranged between 65.1% and 86.8% (Gwet's AC1 0.30 to 0.80). There was substantial or almost perfect agreement (Gwet's AC1 >0.60) in seven cytomorphologic features in the classic PTC group, in six features in the NIFTP group, and in five features in the FA group. There were no features with poor agreement (Gwet's AC1<0.0). CONCLUSIONS:The current study supports the reproducibility of our previous findings. The high level of agreement amongst pathologists for these groups, and particularly the NIFTP group, supports the notion that when viewed in combination as a cytologic profile, these cytomorphologic features may assist the cytopathologist in raising the possibility of NIFTP pre-operatively. This can potentially aid clinicians in deciding whether more conservative treatment may be appropriate.