The diagnostic performance of combined conventional cytology with smears and cell block preparation obtained from endoscopic ultrasound-guided fine needle aspiration for intra-abdominal mass lesions.

BACKGROUND/AIM: Endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) is the primary method for tissue acquisition of intra-abdominal masses. However, the main limitation of cytology alone is the lack of tissue architecture and inadequate samples. This study aimed to evaluate the diagnostic performance of combined conventional cytology and cell block preparation obtained from EUS-FNA of intra-abdominal masses without Rapid On-site Evaluation (ROSE).
METHODS: Cytologic smears and cell block slides of 166 patients undergoing EUS-FNA during 2010-2015 were reviewed by an experienced cytopathologist blinded to clinical data.
RESULTS: 125 patients had neoplastic lesions. Pancreatic adenocarcinoma was the most common etiology (35.5%), followed by lymph node metastasis (27.7%). The mean mass size was 2.5±1.3 cm. The mean number of passes was 1.9±1.28. Tissue adequacy for conventional cytology and cell block preparation was 78.9% and 78.1%, respectively. Factors associated with tissue adequacy were assessed. For cytology, lesions of > 2.1 cm, masses in the pancreatic body or tail, malignancy, and pancreatic cancer were positively associated with adequate cellularity. For cell block preparation, lesions of > 3 cm and malignancy were associated with increased tissue adequacy. The conventional cytology alone had a sensitivity of 68.5%, a specificity of 95.7%, and an area under the receiver operating characteristics (AUROC) of 0.821. The cell block preparation alone had a sensitivity of 65.4%, a specificity of 96%, and an AUROC of 0.807. The combined conventional cytology and cell block preparation performed significantly better than either method alone (p<0.05), as demonstrated by an increased AUROC of 0.853. Furthermore, cell block detected malignancy in 9.3% of cases where the cytologic smears failed to identify malignant cells.
CONCLUSIONS: The combined conventional cytology and cell block preparation increased the diagnostic accuracy of EUS-FNA compared to either method alone. This approach should be implemented in routine practice, especially where ROSE is unavailable.

Introduction

The management of patients with intra-abdominal mass lesions can be challenging in clinical practice as cross-sectional imaging alone is not always sufficient to provide the diagnosis, therefore tissue sampling may be required to decide optimal therapeutic options. Traditionally, ultrasound and computed tomography-guided biopsies were used to obtain tissue samples from such lesions [1]. Over the years, endoscopic ultrasound (EUS)-guided fine needle aspiration (EUS-FNA) has evolved and become the primary technique for tissue acquisition of intra-abdominal lesions adjacent to the gastrointestinal tract, including subepithelial masses arising from the gut wall, pancreatic lesions, lymph nodes, mesentery, and masses in other solid organs [2-4]. This technique is preferable compared to percutaneous and surgical guided biopsies due to lower morbidity and mortality, possibly lower risk of tumor seeding, and cost-effectiveness [5].

The diagnostic accuracy of EUS-FNA ranges from 64% to 96% [4, 6–8]. This variability in the diagnostic yield could be attributed to the location of the tissue, level of expertise, and the presence of rapid on-site evaluation (ROSE). Several studies have demonstrated that ROSE improves the tissue adequacy and diagnostic accuracy of EUS-FNA; however, this facility is not always available [9-12]. The need for a cytotechnician or cytopathologist to be present in the room during the endoscopic procedure, the lack of dedicated cytopathologists, the increased workload for cytopathologists, and the cost limit the use of ROSE in many areas. Despite being the primary technique for tissue acquisition of gastrointestinal and pancreatic diseases, the lack of tissue architecture and inadequate samples are the main limitations of this method, thus, diagnosing certain conditions such as lymphoma, neuroendocrine tumor, autoimmune pancreatitis, and gastrointestinal stromal tumor (GIST) can be challenging because tissue architecture and additional immunostaining are often required [2, 13].

Several factors, particularly the newly developed biopsy needles and tissue handling and processing techniques, have been investigated to overcome these limitations. The initial biopsy needle design known as a tru-cut needle is technically challenging and increases the risk of complications [14-16]. In contrast, the recently developed fine needle biopsy (FNB) with a side port or specially designed needle tip for core tissue collection has been shown to improve tissue adequacy for histological analysis and obviate the need for ROSE [17-19]. Nonetheless, the use of FNB may be limited in resource-limited areas due to increased expenses. Cell block preparation obtained during conventional EUS-FNA has been increasingly recommended because it allows histological and immunohistochemical examination, especially when ROSE is not available [20, 21].

The current study was conducted with the primary aim to assess the diagnostic performance of combined conventional cytology with smears and cell block preparation obtained from EUS-FNA in intra-abdominal solid masses and compare it to each method alone.

Materials and methods

Patient population

A retrospective review of the EUS database at a tertiary care center from 2010 to 2015 was conducted. This study was carried out following the Declaration of Helsinki. The study was approved by the Faculty of Medicine Siriraj Hospital Institutional Review Board (protocol number 111/2557). The informed written consent was waived given the retrospective nature of the study. The patients who underwent EUS-FNA for intra-abdominal masses were identified. The inclusion criteria included 1) age > 18 years, 2) intra-abdominal solid masses accessible by EUS, 3) complete medical records with at least 12 months follow-up, 4) available cytologic smears and cell block slides for review. Exclusion criteria included 1) cystic lesions 2) incomplete medical records and cytopathological data. Patient demographics, clinical presentation, EUS findings, FNA results, complications, and clinical courses were reviewed and analyzed.

EUS-FNA technique

EUS was performed by an experienced endoscopist, who performed more than 2000 EUS cases at a tertiary care center, using a linear array echoendoscope (GF-UC140P-AL5/AL10, Olympus Corp., Tokyo, Japan). Fine needle aspiration was done using a 22-gauge needle (EZshot; Olympus Medical, Tokyo, Japan, and Echotip; Cook Medical, IN, U.S.A.). The technique involved localization of the target lesion, doppler evaluation, needle puncture, tissue aspiration, and specimen collection. Once the lesion was identified, a doppler ultrasound was used to evaluate the intervening vessels. After identifying the proper window without intervening vessels in the needle passage, the stylet was slightly withdrawn to sharpen the needle tip, and the lesion was punctured under real-time ultrasound guidance. After puncturing, the stylet was removed, and negative pressure was applied with 5 mL of suction, followed by moving the needle to and fro within the lesion using the fanning technique. The suction was stopped before removing the needle from the lesion. The tissue was retrieved, and the stylet was inserted into the needle until the specimens extruded through the needle tip and placed onto glass slides for visual inspection. A macroscopic examination was then performed to identify a few visible whitish core tissues of any length. If inadequate macroscopic material was observed, repeat pass attempts were performed until visible core tissue fragments were obtained. The specimens from each pass were prepared for both cytologic smears and cell block preparation for histological evaluation. The maximum number of passes was six based on a study by Jhala et al. The authors demonstrated that 90% of adequate samples were obtained within 6 passes, after which there was only a slight increase in obtaining an adequate sample [22].

Cytologic smears and cell block preparation

ROSE was not available in routine practice at our center due to the limited number of dedicated cytopathologists specializing in gastrointestinal and pancreatic diseases. After each pass, the stylet was reinserted into the needle to expel the aspirated material on glass slides. The specimens were smeared onto the slide and immediately fixed with 95% ethyl alcohol solution. The alcohol-fixed slides were stained with Papanicolaou stain for cytological examination. The remaining material was placed into a bottle containing 10% formalin solution for cell block preparation. Formalin-fixed tissues were embedded in paraffin, and then slides were made using the standard technique [23]. The slides were stained with hematoxylin and eosin and reviewed. Depending on morphology and the cytopathologist’s decision, additional immunohistochemical studies were performed on the tissue blocks derived from formalin-fixed paraffin-embedded (FFPE) to diagnose lymphoma, GIST, neuroendocrine tumors, or metastatic cancer of unknown origin. The immunohistochemical stains used to differentiate various carcinomas included AE 1/AE 3, cell adhesion molecule (CAM) 5.2, cluster of differentiation (CD) 45, cytokeratin (CK) 7, CK 20, S-100 protein (S-100), vimentin, desmin, caudal-type homeobox transcription factor 2 (CDX-2), thyroid transcription factor 1 (TTF-1), chromogranin, synaptophysin, hepatocyte paraffin 1 monoclonal antibody (Hep par 1), alpha-fetoprotein (AFP), paired box gene (Pax-8), renal cell carcinoma (RCC), CD 10. The information about antibody clones are provided in S1 Table. To differentiate spindle cell tumors, we used CD 117, discovered on gastrointestinal stromal tumors (DOG-1), CD 34, S100, and desmin. CD 20 was used to diagnose B-cell lymphoma. Other special stains included acid-fast bacilli and congo red. This study retrieved all available cell block prepared from FFPE and 95% alcohol-fixed slides for cytological analysis. Fig 1 demonstrates an EUS image of a pancreatic lesion, specimen handling after aspiration, Papanicolaou smears, and the hematoxylin and eosin stain.

Fig 1

Endoscopic ultrasound-guided fine needle aspiration with tissue processing method.

(A) Endoscopic ultrasound-guided fine needle aspiration of a pancreatic mass. (B) The specimen was placed onto a glass slide and smeared. (C) The specimen was placed into a bottle containing formalin for cell block preparation. (D) In low power magnification, a cytologic smear shows multiple large fragments of malignant cells in a background of necrotic debris. (Papanicolaou stain, magnification 40X). (E) In high power magnification, a cytologic smear shows a three-dimensional cluster of malignant cells with cellular crowding, nuclear pleomorphism, irregular nuclear membrane, and coarsely clumped chromatin in a background of necrotic debris. (Papanicolaou stain, magnification 400X). (F) Cell block preparation shows tissue fragments containing infiltrating poorly formed glandular structures containing dysplastic cells. The malignant cells are surrounded by desmoplastic stroma and blood clots (Hematoxylin & Eosin stain, magnification 40X).

Definition of cytology and cell block interpretation

All samples were reviewed by an experienced GI cytopathologist blinded to clinical data and prior cytological and cell block results. The tissue adequacy and cytology classification were assessed based on the Papanicolaou society of cytopathology system for reporting pancreaticobiliary cytology [24]. The specimens were considered adequate if the acquired material provided an acceptable amount of cell from the target lesion for cytological evaluation and adequate cells and tissue architecture for histological assessment. The diagnoses from cytologic smears and cell block preparation were categorized as unsatisfactory, negative for malignancy, atypia, suspicious for malignancy, and positive for malignancy. The diagnosis of malignancy was made if the cytopathology reported positive or suspicious for malignancy. The reports of negative for malignancy and atypia were categorized as no malignancy [25, 26]. In cases of inadequate tissue, the EUS-FNA was repeated if clinical presentations or radiological imaging were suspicious for malignancy.

Diagnostic criteria

The final diagnosis was established by 1) histology from a surgical specimen, 2) cell block interpretation and immunohistochemical stain from FFPE tissue obtained via EUS-FNA, 3) cytological diagnosis, 4) minimum of 12 months follow-up for clinical evaluation and interval imaging for benign lesions. A patient was finally diagnosed with malignancy if there was (1) evidence of malignancy based on cytologic smears or cell block material obtained via (a) EUS-FNA, (b) surgical pathology, or (2) cytopathological results suspicious for malignancy and clinical courses suggesting malignancy. The clinical evidence of malignancy comprised (a) new radiographic abnormalities, including regional or distant metastasis, mass infiltrating blood vessels, or adjacent organs (b) cancer-related mortality. The diagnosis of benign conditions required negative cytopathological assessment and at least 12 months of clinical and imaging follow-up, demonstrating no progression or resolution of the lesions.

Statistical analysis

The data was analyzed using SPSS version 16 (SPSS, Inc., Chicago, Illinois). Continuous variables were summarized as mean ± standard deviation and categorical variables as percentages. Comparison between groups was performed using the χ2 test for categorical variables and the t-test for continuous variables. The area under the receiver operating characteristics (AUROC) curve was constructed to evaluate the overall accuracy of cytology and cell block preparation and compared techniques with the DeLong test. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) with 95% confidence intervals (95% CI) were determined for each method. We also made the evaluation based on the different possible outcomes indicative of malignancy for each method. Logistic regression analysis was used to determine factors affecting tissue adequacy. The data were presented as an odds ratio (OD) with a 95% CI. A p-value of less than 0.05 was considered statistical significance.

Results

Characteristics of the study population

Three hundred and thirty-five patients underwent EUS-FNA for intra-abdominal mass lesions during the study period. One hundred and forty-five patients were excluded due to cystic lesions. Specimens were not available for review in 24 cases; thus, 166 patients were included in the study. The mean age was 58 ± 14 years (18 to 84 years). Abdominal pain was the most common presentation accounting for 34.9%, followed by weight loss (23.4%), jaundice (12.7%), and abdominal masses (4.8%). Cross-sectional imaging demonstrated intra-abdominal masses in 141 patients (84.9%). Of the 166 patients, 125 (75.3%) had neoplastic lesions, and 41 (24.7%) had benign conditions. The final diagnoses included pancreatic adenocarcinoma (35.5%), metastatic lymph nodes (27.7%), inflammatory/reactive change (24.7%), lymphoma (3.6%), gastrointestinal stromal tumor (3.6%), cholangiocarcinoma (1.8%), malignant intraductal papillary mucinous neoplasms (1.2%), sarcoma (0.6%), neuroendocrine tumor (0.6%), and hepatocellular carcinoma (0.6%), as shown in Table 1.

Table 1

Patient characteristics and the definite diagnosis.

Parameters	Value
Age, year (mean ± SD)	58 ± 14
Sex, n (%)
Male	90 (54.2)
Female	76 (45.8)
Clinical manifestation, n (%)
Abdominal pain	58 (34.9)
Weight loss	39 (23.4)
Jaundice	21 (12.7)
Abdominal mass	8 (4.8)
Abnormal imaging	141 (84.9)
Definite diagnosis, n (%)
Pancreatic adenocarcinoma	59 (35.5)
Metastatic lymph nodes	46 (27.7)
Inflammatory/Reactive change	41 (24.7)
Lymphoma	6 (3.6)
Gastrointestinal stromal tumor	6 (3.6)
Cholangiocarcinoma	3 (1.8)
Malignant IPMN	2 (1.2)
Sarcoma	1 (0.6)
Neuroendocrine tumor	1 (0.6)
Hepatocellular carcinoma	1 (0.6)

NOTE. Data are presented as mean ± standard deviation or the number (%) of patients with a condition.

SD, standard deviation; IPMN, intraductal papillary mucinous neoplasms.

Final diagnosis

Among patients with neoplastic lesions, surgical pathology was available to confirm the diagnoses in 16 patients (12.8%). Surgical pathology revealed pancreatic adenocarcinoma (n = 8), gastrointestinal stromal tumor (n = 3), malignant intraductal papillary mucinous neoplasm (n = 1), distal cholangiocarcinoma (n = 1), lymphoma (n = 1), gastric cancer (n = 1) and metastatic carcinoma (n = 1). EUS-FNA guided cytohistological analysis demonstrated malignancies in 76 patients. Both cytologic smears and cell block preparation detected malignancies in 59 patients, cytologic smears alone in 9 patients, and cell block preparation alone in 8 patients. The final diagnoses based on surgical pathology and EUS-FNA guided cytohistological analysis and clinical courses are shown in Table 2. The specimens from cell block preparation obtained from FFPE allowed ancillary studies, including immunohistochemical and acid-fast bacilli stains in 31 cases (18.7%), providing the diagnosis of tuberculous lymphadenitis (n = 12), lymphoma (n = 5), pancreatic adenocarcinoma (n = 4), mesenchymal tumor (n = 4), neuroendocrine tumor (n = 4), hepatocellular carcinoma (n = 1), and IgG4 associated pancreatitis (n = 1).

Table 2

Final diagnoses based on surgical pathology and cytohistological analysis via EUS-FNA.

Final diagnoses	Surgical pathology	EUS-FNA with cytohistological analysis and clinical courses
Malignancy, n (%)	16 (9.6)	109 (65.7)
Pancreatic adenocarcinoma	8 (4.8)	51 (30.7)
Metastatic carcinoma	1 (0.6)	45 (27.1)
Lymphoma	1 (0.6)	5 (3.0)
Gastrointestinal stromal tumor	3 (1.8)	3 (1.8)
Distal cholangiocarcinoma	1 (0.6)	2 (1.2)
Gastric cancer	1 (0.6)	1 (0.6)
Malignant IPMN	1 (0.6)	1 (0.6)
Pancreatic neuroendocrine tumor	n/a	1 (0.6)
Benign, n (%)	0	41 (24.7)

NOTE. Data are presented as the number (%) of patients with a condition. IPMN, Intraductal Papillary Mucinous Neoplasm.

EUS findings

The mean mass size was 2.5 ± 1.3 cm (range 0.7 to 10 cm); approximately 40% of lesions were smaller than 2 cm, 32% were 2–3 cm, and 28% were larger than 3 cm. The most common site was the pancreas (51.8%), followed by intra-abdominal lymph nodes (37.9%) and bowel wall (6%). Approximately 80% of the lesions were hypoechoic suggestive of solid lesions. The mean number of needle passes was 1.9±1.3 (range, 1–6). The endosonographic characteristics of the lesions are summarized in Table 3. There were no adverse events related to the EUS-FNA.

Table 3

Endosonographic characteristics of intra-abdominal mass lesions.

Characteristics	Value
Size, cm	2.5 ± 1.3
Number of needle passes	1.9 ± 1.3
Echogenicity of the lesion, n (%)
Hypo-echoic	131 (78.9)
Hetero-echoic/mixed	32 (19.3)
Iso-echoic	2 (1.2)
Hyper-echoic	1 (0.6)
Location of the lesion, n (%)
Pancreas	86 (51.8)
• Head	49 (56.9)
• Body	26 (30.2)
• Tail	11 (12.7)
Intra-abdominal lymph node	63 (37.9)
Bowel wall	10 (6.0)
Liver	4 (2.4)
Retroperitoneal mass	3 (1.8)

NOTE. Data are presented as mean ± standard deviation or the number (%) of patients with a condition.

Tissue adequacy

The percentage of tissue adequacy for cytology and cell block preparation was 78.9% and 78.1%, respectively. Of the 35 patients who had inadequate cellularity for cytologic smears, 6 patients (17.1%) had adequate specimens for cell block preparation. Factors associated with tissue adequacy were location, size, and the nature of lesions, as shown in Table 4. For cytology, lesions of > 2.1–3 cm, masses located in the pancreatic body and tail, malignancy, and pancreatic cancer were positively associated with adequate cellularity. In contrast, intra-abdominal lymph nodes and inflammatory lesions (e.g., chronic pancreatitis and tuberculosis) were negative predictors for tissue adequacy. For cell block preparation, lesions > 3 cm and malignancy were associated with increased tissue adequacy, whereas inflammatory lesions were associated with decreased tissue adequacy.

Table 4

Factors associated with tissue adequacy of specimens obtained by EUS-FNA.

Factor	Cytology				Cell block preparation
	Adequacy (n = 131)	Inadequacy (n = 35)	Unadjusted OD (95% CI)	P-value	Adequacy (n = 125)	Inadequacy (n = 41)	Unadjusted OD (95% CI)	P-value
Number of needle passes, median (IQR)	2 (1–3)	1 (1–3)	1.27 (0.94–1.72)	0.116	1 (1–3)	1 (1–3)	1.11 (0.84–1.46)	0.477
Pancreatic lesion, n (%)
Head	39 (29.8)	10 (28.6)	1.06 (0.47–2.41)	0.890	37 (29.6)	12 (29.3)	1.02 (0.47–2.20)	0.968
Body/Tail	36 (27.5)	1 (2.9)	12.9 (1.70–97.6)	0.013	31 (24.8)	6 (14.6)	1.92 (0.74–5.01)	0.180
Intra-abdominal lymph nodes, n (%)	42 (32.1)	21 (60.0)	0.31 (0.15–0.68)	0.003	44 (35.2)	19 (46.3)	0.63 (0.31–1.29)	0.204
Size of the lesion, n (%)
<2 cm	25 (19.1)	17 (48.6)	1 (Reference)		25 (20.0)	17 (41.5)	1 (Reference)
2.1–3.0 cm	38 (29.0)	11 (31.4)	2.55 (1.02–6.40)	0.045	38 (30.4)	11 (26.8)	2.35 (0.94–5.84)	0.066
>3.0 cm	68 (51.9)	7 (20.0)	7.18 (2.64–19.5)	<0.001	62 (49.6)	13 (31.7)	3.24 (1.37–7.65)	0.007
Echogenicity of the lesion, n (%)
Hypo-echoic	103 (78.6)	28 (80.0)	0.58 (0.20–1.63)	0.299	97 (77.6)	34 (82.9)	0.71 (0.29–1.78)	0.470
Hetero-echoic	27 (20.6)	5 (14.3)	1.56 (0.55–4.39)	0.402	26 (20.8)	6 (14.6)	1.53 (0.58–4.03)	0.388
Final diagnosis, n (%)
Malignancy	108 (82.4)	17 (48.6)	4.97 (2.23–11.1)	<0.001	100 (80.0)	25 (61.0)	2.56 (1.19–5.50)	0.016
Pancreatic cancer	57 (43.5)	4 (11.4)	10.8 (3.28–35.3)	<0.001	51 (40.8)	10 (24.4)	2.14 (0.96–4.74)	0.062
Lymphoma	5 (3.8)	1 (2.9)	3.91 (0.42–36.5)	0.231	5 (4.0)	1 (2.4)	1.67 (0.19–14.7)	0.646
Inflammatory disease	23 (17.6)	18 (51.4)	0.20 (0.09–0.45)	<0.001	25 (20.0)	16 (39.0)	0.39 (0.18–0.84)	0.016

CI, confidence interval; IQR, interquartile range; OD, odds ratio.

Diagnostic performance of cytology and cell block preparation

The performance of cytology and cell block preparation for diagnosing malignancy are shown in Table 5. Overall, cytology had an intermediate sensitivity of 53.7% but a high specificity of 95.7% for diagnosing cancer when only cytological diagnosis classified as positive for malignancy was considered positive. It is noteworthy that cytological analysis categorized as suspicious for malignancy had a specificity and PPV of 100%; therefore, it should be considered equivalent to that classified as positive for malignancy. If both suspicious and positive for malignancy were considered positive, the sensitivity increased to 68.5%, and the specificity remained at 95.7%. Similarly, cell block preparation had an intermediate sensitivity of 57.4% and a high specificity of 95.7% for diagnosing malignancy when only positive for malignancy was considered positive. When only suspicious results were analyzed, the sensitivity was as low as 7.92%, whereas the specificity was 100%. The sensitivity of cell block preparation increased to 65.4% when both positive and suspicious results were considered positive. However, the diagnostic performance of cell block preparation and cytology for diagnosing malignancy were comparable (the AUROCs were 0.807 vs. 0.821, respectively; p = 0.618). For combined cytology and cell block preparation, when both positive and suspicious results of either test were considered positive, the sensitivity increased significantly from 64.2% to 74.6%, and the specificity remained high (96%). The AUROC of combined cytology and cell block preparation for diagnosing cancer was 0.853, and this was significantly better than either cytology (p = 0.002) or cell block preparation (p = 0.006) alone.

Table 5

Diagnostic performance of cytology and cell block preparation obtained by EUS-FNA of intra-abdominal masses.

	AUROC (95%CI)	Sensitivity (%) (95%CI)	Specificity (%) (95%CI)	PPV (%) (95%CI)	NPV (%) (95%CI)
Cytology
Positive or suspicious	0.821	68.5	95.7	98.7	39.3
	(0.744–0.882)	(58.9–77.1)	(78.1–99.1)	(91.6–99.8)	(32.6–46.4)
Positive	0.747	53.7	95.7	98.3	30.6
	(0.663–0.819)	(48.3–63.3)	(78.1–99.9)	(89.4–99.7)	(26.1–35.4)
Suspicious	0.574	14.8	100	100	20.0
	(0.485–0.660)	(8.7–22.9)	(85.2–100)		(18.8–21.3)
Cell block preparation
Positive or suspicious	0.807	65.4	96.0	98.5	40.7
	(0.727–0.872)	(55.2–74.5)	(79.6–99.9)	(90.6–99.8)	(34.1–47.6)
Positive	0.767	57.4	96.0	98.3	35.8
	(0.684–0.838)	(47.2–67.2)	(79.6–99.9)	(89.4–99.7)	(30.5–41.5)
Suspicious	0.540	7.92	100	100	21.2
	(0.449–0.629)	(3.5–15.0)	(86.3–100)		(20.2–22.2)
Cytology and cell block preparation
Either test with	0.853	74.6	96.0	98.8	46.2
positive or suspicious	(0.782–0.908)	(65.4–82.4)	(79.6–99.9)	(92.3–99.8)	(38.1–54.4)
Either test with	0.801	64.2	96.0	98.6	38.1
positive	(0.723–0.865)	(54.5–73.2)	(79.6–99.9)	(91.1–99.8)	(32.1–44.5)
Either test with	0.586	17.3	100	100	21.6
suspicious	(0.498–0.670)	(10.7–25.7)	(86.3–100)		(20.1–23.0)

PPV, positive predictive value; NPV, negative predictive value; CI, confidence interval; AUROC, area under the Receiver-Operating Characteristic Curve.

Furthermore, we explored whether cell block preparation would help detect malignancy when cytology was inadequate or unsuccessful. Of 86 patients who had either inadequate cytology or cytological diagnosis of negative for malignancy, 58 patients had adequate specimens for cell block preparation. Of these, 8 patients (9.3%) were found to have malignancy, including pancreatic and metastatic cancer. Other diagnoses included tuberculous lymphadenitis (n = 10), reactive lymph nodes (n = 9), abscess (n = 3), GIST (n = 1), benign pancreatic tissue (n = 1). The diagnoses could not be made by cell block evaluation in 26 patients. Among these, malignancy (n = 16), tuberculosis (n = 6), and other benign conditions (n = 4) were discovered during clinical and imaging follow-up.

Discussion

Despite being a primary modality for tissue acquisition for intra-abdominal solid lesions, EUS-FNA carries certain limitations, such as the inability to obtain core tissue for histological features and inadequate sampling. Cell block preparation has been recognized as a powerful technique for assessing tissue architecture and determining its histological features. Results from retrospective studies suggest that histological evaluation from cell block preparation increases the diagnostic accuracy of EUS-FNA for malignancy [20, 21, 27–29]. The European Society of Gastrointestinal Endoscopy (ESGE), based on low-quality evidence, recommends that EUS-guided tissue sampling should include histologic preparations (e.g., cell blocks) and not be limited to cytology [30]. Therefore, we conducted a comparative analysis of the diagnostic performance between cytology vs. cell block preparation for histological evaluation vs. combined methods in the absence of ROSE.

The present study showed that cytology alone had a sensitivity of 68.5%, a specificity of 95.7%, and an AUROC of 0.821. Cell block preparation alone had a sensitivity of 65.4%, a specificity of 96%, and an AUROC of 0.807. Cytology and cell block preparation combined performed significantly better than either method alone with an AUROC of 0.853. Also, cell block preparation could detect malignancy in patients who had inadequate specimens for cytology or cytological diagnosis classified as negative for malignancy. These findings emphasize the importance of both cytologic smears and cell block preparation during tissue processing.

In the absence of ROSE, the tissue adequacy for cytological diagnosis may decrease by 10–15% [31]. The reported specimen adequacy ranges from 70%-92% for cytology and 68%-86.5% for cell block preparation for histology using different needle sizes (18G, 22G, 25G) [20, 21, 32]. The present study demonstrated similar results with 78.9% adequacy for cytology and 78.1% for cell block preparation. The factors influencing tissue adequacy have been explored. We found that pancreatic body and tail lesions, tumor size of >2 cm, the diagnosis of malignancy, and pancreatic cancer had a positive association with tissue adequacy, whereas lesions of lymph nodes and inflammatory diseases had a negative association with tissue adequacy for cytological evaluation. For cell block preparation, only tumor size of >30 mm and malignancy were associated with increased tissue adequacy. In contrast, inflammatory diseases were associated with decreased tissue adequacy. It is worth mentioning that pancreatic body and tail lesions were associated with increased tissue adequacy for cytological analysis but pancreatic head lesions were not. This may be related to the technical aspect of EUS-FNA. Lesions in the body and tail are generally easier to puncture because the echoendoscope is in a straight position allowing the needle to puncture through easily compared to the head lesions.

In terms of the number of needle passes, the ESGE recommends 3–4 passes in the absence of ROSE (30). However, the present study demonstrated the mean number of passes of 1.9, which is similar to Moeller et al. In their study, the mean number of passes was 1.88 with the diagnostic adequacy of 98.9%, and the sensitivity of 82.9% [21]. We hypothesize that the mean number of passes in our study is lower than the ESGE recommendation because a macroscopic evaluation was performed after each pass, allowing the endosonographer to stop the tissue aspiration once the fragments of tissue core were observed. Recently, the macroscopic on-site evaluation (MOSE) of the aspirates has been introduced; however, the method has not been standardized. A macroscopic visible core of ≥ 4 mm in length is associated with a higher diagnostic yield [33]. A randomized trial comparing the diagnostic yield of MOSE to conventional EUS-FNA without ROSE has shown that both techniques provide a similar yield, but MOSE requires fewer passes than the conventional method [34]. These findings demonstrate the diagnostic value of MOSE, especially in the absence of ROSE. Our study performed an onsite macroscopic examination; nonetheless, the criteria were different from MOSE described in the recent literature. We accepted any length of a few visible cores.

There is a variety of specimen handling and processing methods in clinical practice. We used the specimens obtained from each pass for cytologic smears and cell block preparation. Therefore, no additional passes or new needles were required to obtain tissue for the cell block preparation. Direct smears are often performed, and alcohol or saline is used for liquid-based cytology. Formalin is commonly used for histology preservation [30]. Also, cell block preparation using the sodium alginate method, stained with hematoxylin and eosin, and immunohistochemistry have been described [35]. The refinement and standardization of the tissue processing techniques may enhance the diagnostic accuracy of EUS-FNA and deserve further investigation.

Another essential factor influencing the diagnostic performance of cytology and cell block is the categorization and interpretation of specimens in correlation with the clinical course. We proposed that cytological and cell block with histological assessment classified as suspicious for malignancy should be considered positive because of its excellent specificity and PPV of 100%. In addition, clinical courses should be taken into consideration when malignancy is suspected on cytological or histological evaluation.

The strengths of this study included 1) large sample size, 2) the results can be applied to small (<2 cm) and large (>3 cm) lesions, 3) a dedicated GI cytopathologist, blinded to clinical data, prospectively reviewed both the cytologic smears and slides from cell block preparation to avoid interpretation bias, and 4) a complete clinical data and long-term follow-up period. However, this was a single-center, single endoscopist-based, single cytopathologist, and a retrospective review. A vast majority of patients did not have surgical pathology. Hence, prospective comparative studies or randomized, controlled trials may be warranted.

Conclusion

The combination of cytology and cell block preparation improves the diagnostic performance of EUS-FNA in intra-abdominal solid masses compared to either method alone. This approach shows promise and should be routinely implemented in clinical practice, especially where on-site cytopathology is unavailable.

The antibody clones for immunohistochemistry staining.

(DOCX)

Click here for additional data file.

(PDF)

Click here for additional data file.

27 Oct 2021

11 Dec 2021

Dear Editor,

We would like to thank the editor and both reviewers for the constructive criticisms, which have helped us improve our manuscript. As suggested, we have revised the manuscript point-by-point in response to the reviewers’ questions.

Reviewers' comments:

Reviewer #1: The paper carries out a comparative analysis of EUS-FNA cytology and histology diagnostic performances in the context of solid intra-abdominal mass lesions, using a large sample size. The factors that positively and negatively correlate with the adequacy of the samples are identified. The two compared methods did not demonstrate different diagnostic performances, but their combination ensured an AUROC significantly better than either method alone.

Response: We thank the reviewer for taking the time to review our manuscript and providing helpful input to improve the clarity of our manuscript. We have revised our manuscript to address the reviewer’s comments and provide a point-by-point response to the specific questions.

Major issues:

1) The whole DISCUSSION (starting from line 250) is quite confusing and needs an overall review.

Response: The entire discussion section has been revised to improve the clarity of the manuscript, as suggested by the reviewer. In brief, the first paragraph stated the role of EUS-FNA in the diagnosis of intra-abdominal lesions and the limitations of the test. The second paragraph presented the study's main results, including the diagnostic performance of cytological analysis alone, histological analysis alone, and combined method. The third paragraph discussed the percentage of tissue adequacy from our study compared to the literature and the factors associated with tissue adequacy. The fourth paragraph touched on the number of needle passes and the usefulness of macroscopic onsite evaluation of specimens, especially in the absence of ROSE. The fifth paragraph described various tissue handling and processing methods and how the specimens were processed in our study. The discussion in the sixth paragraph focused on how to interpret cytohistological specimens based on their category in correlation with clinical courses. Lastly, the strengths and weaknesses of the study were stated in the last paragraph.

Minor issues:

1) “Supporting info data” layout has some problems, the last column is moved at the end of the document.

Response: The PDF file of the “supporting info data” has been reformatted and all columns are shown properly.

2) Figure 1: the quality is poor and the boxes should be properly framed.

Response: We have revised Figure 1 to improve the image quality. The boxes are properly framed. Also, we have added images of cytology smears (PAP smear) to compare with the histology slide (hematoxylin and eosin stain) derived from cell block preparation.

3) An abbreviations section is needed.

Response: An abbreviations section has been added after the abstract.

4) Lines 56-58: “The management of patients with intra-abdominal mass lesions can be challenging in clinical practice as cross sectional imaging alone is not always sufficient to provide the diagnosis, therefore tissue sampling may be required to decide optimal therapeutic options.”

Response: The changes have been made as suggested.

5) Line 59: tissue samples.

Response: “tissue diagnosis” has been changed to “tissue samples” as suggested.

6) Lines 68-70: “The lack of tissue architecture and inadequate samples are the main limitations of this method, thus diagnosing certain conditions… [1]”

Response: The changes have been made as suggested.

7) Line 72: the new line should be used here and not in line 74.

Response: The change has been made as suggested.

8) Line 75: increases.

Response: The change has been made as suggested.

9) Patient population: lines 162-164 should be moved here.

Response: The sentences from line 162-164 have been moved to “patient population” under Materials and Methods section as suggested.

10) Line 88: why did your analysis consider only the years 2010-2015?

Response: Our center started a formal EUS program in 2009. Initially, FNA needles were used in all cases. The trend started to change after Fine needle biopsy (FNB) was introduced to our center in 2016. A vast majority of patients have been included in several FNB-related projects. Therefore, we elected to recruit patients between 2010-2015.

11) Line 132: from a qualitative point of view, none of the 166 cases had any problems?

Response: Besides overall tissue adequacy being 78%, some minor problems were air artifacts and blood contamination on the cytology smears. However, these issues did not significantly interfere with the interpretation of the specimens. We routinely prepared 4-6 slides for cytology. Air artifacts may have occurred in some slides but not in all. We avoided using suction during tissue aspiration for hypervascular lesions to minimize blood contamination. Also, a doppler was used to detect intervening vessels before puncturing in all lesions, as mentioned in the manuscript.

12) Line 137: rephrase.

Response: The sentence was rephrased to “In cases of inadequate tissue, the EUS-FNA was repeated if clinical presentations or radiological imaging were suspicious for malignancy.”

13) Line 141: it should be said also here that the point 4) is referred just to benign lesions.

Response: We have clarified that the point 4) is referred benign lesions by adding “for benign lesions” as suggested.

14) Line 141: “A patient was finally diagnosed…”

Response: The change has been made as suggested.

15) Lines 169-172: this means that the study of the cytology + histology combination was based just on 69 cases? If so, it should be stated.

Response: The combination of cytological and histological analysis was performed in all recruited patients (N=166). However, 69 patients had malignant cells detected by both cytology smears and cell block preparation for histological evaluation. Fourteen patients had malignant cells detected by cytological evaluation alone, and 12 patients had malignant cells detected by histological analysis alone. Surgical pathology was available to confirm the diagnoses in 16 patients. We have clarified this issue by adding the subheading “final diagnosis” with a detailed explanation and a table summarizing the explanation (Table 2) to the Results section. The order of the following tables has been changed accordingly.

16) Lines 172 and 173: gastrointestinal and gastric.

Response: Changes have been made as suggested.

17) Lines 175: delete comma after bracket.

Response: Comma after bracket was deleted.

18) Table 1: definite diagnosis sum is 164, are 2 cases missing?

Response: Table 1 was revised (N=166). Two missing cases were malignant intraductal papillary mucinous neoplasms (IPMN).

19) Table 2: Hetero-echoic/mixed

Response: The terminology has been modified (mixed was added) as suggested.

20) Table 3: specify the meaning of OD. A visual way to point out if a factor has a positive or a negative correlation should be found.

Response: To point out whether a factor has a positive or negative correlation with tissue adequacy, we added detailed data demonstrating the probability of each factor to be associated with tissue adequacy. The relative probability of each factor was then presented as the odds ratio. Table 3 (the table's order was changed to 4 after revision) has been revised.

21) Line 212: “... was considered actually positive.”

Response: The change has been made as suggested. “actually” was added.

22) Line 221-223: it should be said if there is significance for Sn and Sp alone.

Response: When both positive and suspicious results of either test were considered positive, the cytological and histological analysis combined significantly increased the sensitivity from 64.2 to 74.6% (p< 0.05) and the specificity remained unchanged at 96%. The change was made by stating that the sensitivity increased significantly.

23) Lines 241-243: “The European Society of Gastrointestinal Endoscopy (ESGE), based on low-quality evidence, recommends that EUS-guided tissue sampling should include histologic preparations (e.g., cell blocks) and not be limited to cytology [28].”

Response: Changes have been made as suggested.

24) Line 293: a single cytopathologist is another limitation.

Response: A single cytopathologist was added to the limitations as suggested.

Reviewer #2: This research article focuses on the cytological evaluation of intra-abdominal mass lesions. In particular, it is a retrospective study comparing the diagnostic performance of samples obtained through endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) and processed in different ways (conventional cytology with smears versus cell block preparation for histology versus the two methods combined) in a setting that does not allow the use of Rapid On-Site Evaluation (ROSE); it also investigates the different factors that affect the adequacy of the cytological sampling through EUS-FNA.

The study focuses on the evaluation of a sampling method which leads to obtaining cytological samples, which are considered small samples, and highlights their diagnostic strength, especially in case of a combination of conventional cytology with smears and cell block preparation for histology. It is well constructed, the discussion is well articulated, well written and fluid and the strengths and limits of the study are clearly defined.

However, there are some issues that need to be addressed:

Response: We thank the reviewer for taking the time to review our manuscript and providing helpful comments to improve the quality of our manuscript. We have revised our manuscript to address the reviewer’s comments and a point-by-point response to the specific questions are provided.

MAJOR ISSUES

-The study compares cytological samples obtained through EUS-FNA and processed in two different ways, in particular conventional cytology with smears and cell block preparation for histology. In the title and in the text, however, the word "histology" is often used directly in place of “cell block preparation”. In my opinion, the terminology must be revised, highlighting the fact that your group is focusing on cytological samples, which then, in the case of the cell block preparations, are also able to provide further information (e.g. tissue architecture, possibility of immunohistochemical examination) in comparison to conventional cytological smears.

Response: The reviewer’s comments are well taken. As the reviewer suggested, we have made changes to the terminologies related to cell block preparation and histology throughout the manuscript. The term “histology” was modified to “cell block preparation for histological analysis”. Throughout the article, we were more consistent with the terms/descriptions of tissue processing methods. Also, the manuscript title has been changed to “The diagnostic performance of combined conventional cytology with smears and cell block preparation for histological evaluation obtained from endoscopic ultrasound-guided fine needle aspiration for intra-abdominal mass lesions”.

- It is clear from the beginning that ROSE was not used. Since ROSE is one of those elements which more affects the adequacy of the EUS-FNA, it would be useful, starting from the introduction, to explain the reasons why ROSE is not always available and also later, in the section “Materials and methods”, the reasons why it was not possible to use it in your study.

Response: We have added the following reasons to the introduction to explain why ROSE is not always available. "The need for a cytotechnician or cytopathologist to be present in the room during the endoscopic procedure, the lack of dedicated cytopathologists, the increased workload for cytopathologists, and the cost limit the use of ROSE in many areas." These sentences have been highlighted. Regarding our center, ROSE service could not be offered in routine practice because of the lack of human resources and a limited number of dedicated GI cytopathologists. This has been added to the "cytology and histology preparation" under the materials and methods section.

-Fig. 1: you should specify the magnification of the hematoxylin and eosin picture. If possible, you should also add the conventional cytology counterpart.

Response: Figure 1 has been revised. The magnification of each picture has been added. Images of cytologic smears for PAP stain have been added to compare with the histology slide (H&E stain) derived from cell block preparation.

-Page 12, lines 254-257: you should provide a possible explanation of the low number of needle passes obtained, in comparison to the European Society of Gastrointestinal Endoscopy (ESGE) recommendations.

Response: We hypothesize that the mean number of passes in our study is lower than the ESGE recommendation because a macroscopic evaluation was performed after each pass, allowing the endosonographer to stop the tissue aspiration once the fragments of tissue core were observed. This possible explanation has been added to the discussion.

-Discussion: you should provide a possible explanation of how the location of the lesion affects the adequacy of the sampling.

Response: The factors affecting tissue adequacy have been elaborated in more detail, and a possible explanation of how the location of the lesion affects the adequacy of the sampling has been added to the discussion. The following paragraph has been added. “The factors influencing tissue adequacy have been explored. We found that pancreatic body and tail lesions, tumor size of >2 cm, the diagnosis of malignancy, and pancreatic cancer had a positive association with tissue adequacy, whereas lesions of lymph nodes and inflammatory diseases had a negative association with tissue adequacy for cytological evaluation. For histological evaluation, only tumor size of >30 mm and malignancy were associated with increased tissue adequacy. In contrast, inflammatory diseases were associated with decreased tissue adequacy. It is worth mentioning that pancreatic body and tail lesions were associated with increased tissue adequacy for cytological analysis but pancreatic head lesions were not. This may be related to the technical aspect of EUS-FNA. Lesions in the body and tail are generally easier to puncture because the echoendoscope is in a straight position allowing the needle to puncture through easily compared to the head lesions.”

MINOR ISSUES

-Page 3, line 59: you should add some references from literature regarding traditional methods used to obtain tissue diagnosis from abdominal mass lesions.

Response: A reference has been added as suggested. Lipnik AJ, Brown DB. Image-Guided Percutaneous Abdominal Mass Biopsy: Technical and Clinical Considerations. Radiol Clin North Am. 2015;53(5):1049-59.

-Page 4, line 106-109: you should clarify the criteria used for the macroscopic evaluation/visual inspection of the specimens obtained from EUS-FNA in your study.

Response: The criteria used for the macroscopic evaluation have been added. The visual inspection should identify a few visible whitish core tissues of any length.

-Page 4, line 109: does the number "six passes" derive from your retrospective experience? Was it based on some literature data?

Response: The number “six passes” is based on a study by Jhala et al. (Jhala NC, Jhala D, Eltoum I, Vickers SM, Wilcox CM, Chhieng DC, et al. Endoscopic ultrasound-guided fine-needle aspiration biopsy: A powerful tool to obtain samples from small lesions. Cancer cytopathology. 2004;102(4):239-46)

The following explanation has been added. The maximum number of passes was six based on a study by Jhala at el. The authors demonstrated that 90% of adequate samples were obtained within 6 passes, after which there was only a slight increase in obtaining adequate sample (22).

-Page 5, line 117: you should provide a list of the antibodies used for the additional immunohistochemical studies.

Response: The list of antibodies has been added. The immunohistochemical stains used to differentiate various carcinomas included AE 1/3, CAM 5.2, CD 45, CK7, CK 17, CK 20, S-100, vimentin, Desmin, CDX-2, TTF-1, chromogranin, synaptophysin, Hep par 1, AFP, Pax-8, RCC, CD 10, cytokeratin. For spindle cell tumor, we used CD 117, DOG-1, CD 34, S100 protein, and Desmin. CD 20 was used to diagnose B-cell lymphoma. Other special stains included acid-fast bacilli and congo red.

-Page 5, line 137: "Repeat tissue acquisition may be performed as indicated": what does this sentence mean?

Response: This sentence has been rephrased to “In cases of inadequate tissue, the EUS-FNA was repeated if clinical presentations or radiological imaging were suspicious for malignancy.”

-Page 7, lines 168-180: in addition to the data in the text, it would be useful to create a table with the diagnosis (malignant and benign) and where they derive from (smears, cell block preparations, surgical pathology); also you should move the paragraph regarding the immunohistochemistry to the end of the section.

Response: A new table has been created (table 2) as the reviewer suggested. Also, we added the subheading “final diagnosis” with detailed explanation about how the diagnoses were derived along with the new table2 summarizing the findings. The paragraph regarding the immunohistochemistry is included in this section.

-Page 7, lines 162-180: you should report in the text the same percentages reported in Table 1.

Response: Changes have been made as the reviewer suggested.

-Page 7, lines 178-180: these criteria were already specified before, in the “Material and methods” section.

Response: These sentences have been removed.

-Pages 8-9, table 2: you should add in brackets the explanation of the values for "Size, cm" and "Number of needle passes'', as in table 1 ("mean +/- SD")

Response: Changes have been made.

-Pages 8-9, table 3: you should explain the abbreviations "OD" and "CI"; what does "Reference" mean regarding the "Size of the lesion"?

Response: Explanation of the abbreviations have been added below the table. Regarding “reference”, it indicates that the lesion size of 2 cm was used as the reference (value =1) for comparison. The changes have been made and highlighted.

Submitted filename: Response to reviewers.docx

Click here for additional data file.

7 Jan 2022

30 Jan 2022

Dear Editor,

We would like to thank the editor and both reviewers for reviewing our manuscript and providing constructive criticisms. As suggested, we have revised the manuscript point-by-point in response to the reviewers’ questions.

Reviewers' comments:

Reviewer #1: The authors have assessed the issues clarifying doubts and appropriately modifying the paper, therefore in my opinion the article can be submitted for publication.

Response: We are grateful for the reviewer’s input which helps improve the quality of our manuscript. Your time is greatly appreciated.

Reviewer #2: I would like to thank the authors for all the answers to my comments and the revision of the manuscript. In my opinion there are still some issues that need to be addressed before publication:

Response: We thank the reviewer for taking the time to review our manuscript and providing helpful comments. We have revised our manuscript to address the reviewer’s comments, and a point-by-point response to the specific questions is provided.

-The word "histology" is still often used directly in place of “cell block preparation”. In my opinion, the terminology must be checked through all the text (considering also the abstract and the tables).

Response: Thank you for pointing this out. To make the terminology consistent throughout the article, we have changed the terminology regarding the tissue processing method (cytology, cytological evaluation, histology, cell block preparation) throughout the manuscript. Changes have been made to all sections, including the manuscript title, abstract, introduction, materials and methods, results, discussion, conclusion, tables, and the figure legend.

1. The manuscript title has been changed to “The diagnostic performance of combined conventional cytology with smears and cell block preparation obtained from endoscopic ultrasound-guided fine needle aspiration for intra-abdominal mass lesions”.

2. The word “cytological analysis” was replaced by “cytology” or “conventional cytology” or “cytologic smears” when appropriate.

3. The word “histology” was replaced by “cell block preparation” or “cell block preparation for histological assessment” when appropriate.

-Line 31 and others: ”combined cytological”: you should change it to “cytology” or “conventional cytology” through all the text.

Response: Changes have been made as suggested. “combined cytological….” was changed to “combined conventional cytology and cell block preparation…..”

-Lines 127-130: “Three hundred and thirty-five patients underwent EUS-FNA for intra-abdominal mass lesions during the study period. One hundred and forty-five patients were excluded due to cystic lesions. Specimens were not available for review in 24 cases; thus, 166 patients were included in the study.” : you should report this part in the result section.

Response: Changes have been made as suggested. Lines 127-130 were moved to the result section.

-Line 150: “ROSE is not available”: “ROSE was not available”.

Response: Change has been made as suggested.

-Lines 159-166: “The immunohistochemical stains…”: you should mention the clones you used.

Response: A summary of antibodies and clones used in this study has been put together in Supplementary table 1 (please see the attached supplementary table 1). The following sentence, “The information about antibody clones are provided in Supplementary Table 1” was added to the text under the Materials & Methods section, subheading “Cytological smears and cell block preparation”.

-Line 178 “(Pap)”: in my opinion you should use either the entire word or the abbreviation.

Response: Pap was changed to Papanicolaou.

-Table 2, “Surgical pathology” column: you should check the concordance with the data reported in the text.

Response: The concordance between the data in table 2 and the text was reviewed. Table 2 has been revised as suggested.

-Table 3: you should delete the “%” signs.

Response: “%” signs were removed.

-Table 4: you should delete the word “abbreviation”.

Response: The word “abbreviation” below Table 4 was removed.

-Line 290: where does the data “13 patients” come from?

Response: We intended to state that in cases of surgical pathology proven malignancy, the diagnostic accuracy of EUS-FNA was 100%. The number of patients with surgical pathology was 16, not 13. “13” was a typo. However, we elected to remove this sentence as the number of patients with surgical pathology proven malignancy was quite limited, as mentioned in the study limitations, and added limited value.

-Lines 292-293, “Of 86 patients without the diagnosis of malignancy based on cytological evaluation (inadequate = 28, negative for malignancy = 58)”: you should report in the text the results of the FNAB of all the patients in this study.

Response: Of 86 patients who had either inadequate cytology or cytological diagnosis of negative for malignancy, 58 patients had adequate specimens for cell block preparation. Of these, 8 patients (9.3%) were found to have malignancy, including pancreatic and metastatic cancer. Other diagnoses included tuberculous lymphadenitis (n = 10), reactive lymph nodes (n = 9), abscess (n = 3), GIST (n = 1), benign pancreatic tissue (n = 1). The diagnoses could not be made by cell block evaluation in 26 patients. Among these, malignancy (n = 16), tuberculosis (n = 6), and other benign conditions (n = 4) were discovered during clinical and imaging follow-up. These results have been added as suggested.

Submitted filename: Response to reviewers.RV2.docx

Click here for additional data file.

2 Feb 2022

The diagnostic performance of combined conventional cytology with smears and cell block preparation obtained from endoscopic ultrasound-guided fine needle aspiration for intra-abdominal mass lesions

PONE-D-21-31814R2

Dear Dr. Pausawasdi,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Vincenzo L'Imperio

Academic Editor

PLOS ONE

14 Mar 2022

PONE-D-21-31814R2

The diagnostic performance of combined conventional cytology with smears and cell block preparation obtained from endoscopic ultrasound-guided fine needle aspiration for intra-abdominal mass lesions

Dear Dr. Pausawasdi:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Vincenzo L'Imperio

Academic Editor

PLOS ONE