Advertisement

Comparison of liquid-based cytology with conventional smear cytology for EUS-guided FNA of solid pancreatic masses: a prospective randomized noninferiority study

  • Author Footnotes
    ∗ Drs Chun and K. Lee contributed equally to this work as first authors.
    Jung Won Chun
    Footnotes
    ∗ Drs Chun and K. Lee contributed equally to this work as first authors.
    Affiliations
    Department of Internal Medicine and Liver Research Institute, Seoul National University Hospital, Seoul National University College of Medicine, Seoul, Korea
    Search for articles by this author
  • Author Footnotes
    ∗ Drs Chun and K. Lee contributed equally to this work as first authors.
    Kyoungbun Lee
    Footnotes
    ∗ Drs Chun and K. Lee contributed equally to this work as first authors.
    Affiliations
    Department of Pathology, Seoul National University Hospital, Seoul National University College of Medicine, Seoul, Korea
    Search for articles by this author
  • Author Footnotes
    † Drs S.H. Lee and H. Kim contributed equally to this work as corresponding authors.
    Sang Hyub Lee
    Correspondence
    Reprint requests: Sang Hyub Lee, MD, PhD, Department of Internal Medicine, Seoul National University Hospital, Seoul National University College of Medicine, 101 Daehak-ro, Jongno-gu, Seoul 03080, Republic of Korea
    Footnotes
    † Drs S.H. Lee and H. Kim contributed equally to this work as corresponding authors.
    Affiliations
    Department of Internal Medicine and Liver Research Institute, Seoul National University Hospital, Seoul National University College of Medicine, Seoul, Korea
    Search for articles by this author
  • Author Footnotes
    † Drs S.H. Lee and H. Kim contributed equally to this work as corresponding authors.
    Haeryoung Kim
    Correspondence
    Haeryoung Kim, MD, PhD, Department of Pathology, Seoul National University Hospital, Seoul National University College of Medicine, Basic Science Building 101, 103 Daehak-ro, Jongno-gu, Seoul 03080, Republic of Korea.
    Footnotes
    † Drs S.H. Lee and H. Kim contributed equally to this work as corresponding authors.
    Affiliations
    Department of Pathology, Seoul National University Hospital, Seoul National University College of Medicine, Seoul, Korea
    Search for articles by this author
  • Min Su You
    Affiliations
    Department of Internal Medicine and Liver Research Institute, Seoul National University Hospital, Seoul National University College of Medicine, Seoul, Korea
    Search for articles by this author
  • Yoon Jung Hwang
    Affiliations
    Department of Pathology, Seoul National University Hospital, Seoul National University College of Medicine, Seoul, Korea
    Search for articles by this author
  • Woo Hyun Paik
    Affiliations
    Department of Internal Medicine and Liver Research Institute, Seoul National University Hospital, Seoul National University College of Medicine, Seoul, Korea
    Search for articles by this author
  • Ji Kon Ryu
    Affiliations
    Department of Internal Medicine and Liver Research Institute, Seoul National University Hospital, Seoul National University College of Medicine, Seoul, Korea
    Search for articles by this author
  • Yong-Tae Kim
    Affiliations
    Department of Internal Medicine and Liver Research Institute, Seoul National University Hospital, Seoul National University College of Medicine, Seoul, Korea
    Search for articles by this author
  • Author Footnotes
    ∗ Drs Chun and K. Lee contributed equally to this work as first authors.
    † Drs S.H. Lee and H. Kim contributed equally to this work as corresponding authors.
Open AccessPublished:November 20, 2019DOI:https://doi.org/10.1016/j.gie.2019.11.018

      Background and Aims

      There are limited data on the efficacy of liquid-based cytology (LBC) for EUS-guided FNA specimens. We aimed to evaluate the diagnostic efficacy of LBC for solid pancreatic neoplasms compared with conventional smears (CSs).

      Methods

      In this randomized, crossover, noninferiority trial, we randomly assigned (1:1) patients with suspected pancreatic cancer to the LBC group or the CS group. Aspirates from the first needle pass were processed by one method, aspirates from the second pass by the other method, and specimens from the last pass were processed as core biopsy samples. The primary endpoint was the diagnostic efficacy of each method, with the final diagnosis as the gold standard. A noninferiority margin of −10% was assumed.

      Results

      Of 170 randomized patients, 165 were classified as malignant and 5 as benign. Unsatisfactory samples were less frequent in the LBC group (1.78%) compared with the CS group (5.33%). The diagnostic accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of LBC versus CS were 88.0% versus 83.8% (P = .276), 87.7% versus 83.2% (P = .256), 100% versus 100% (P = .999), 100% versus 100% (P = .999), and 16.7% versus 16.1% (P = .953), respectively. A bloody background was significantly more frequent in the CS group (CS, 85.2%; LBC, 1.8%; P < .001), whereas the nuclear features were similar for both groups.

      Conclusions

      The diagnostic usefulness of LBC was comparable with that of CS. The cytomorphologic features did not differ significantly between the 2 methods, and the reduced bloody backgrounds allowed better visibility in the LBC method. (Clinical trial registration number: NCT03606148.)

      Graphical abstract

      Abbreviations:

      CS (conventional smear), LBC (liquid-based cytology), PanNET (pancreatic neuroendocrine tumor), ROSE (rapid on-site evaluation)

      Introduction

      Pancreatic cancer is the fourth leading cause of cancer-related deaths in the United States, and fifth in South Korea, and its incidence and mortality rates are increasing.
      • Jung K.-W.
      • Won Y.-J.
      • Kong H.-J.
      • et al.
      Cancer statistics in Korea: incidence, mortality, survival, and prevalence in 2015.
      ,
      • Saad A.M.
      • Turk T.
      • Al-Husseini M.J.
      • et al.
      Trends in pancreatic adenocarcinoma incidence and mortality in the United States in the last four decades; a SEER-based study.
      Most patients with pancreatic cancer are not candidates for curative resection at the time of diagnosis and require treatment with systemic chemotherapy.
      • Dusch N.
      • Weiss C.
      • Strobel P.
      • et al.
      Factors predicting long-term survival following pancreatic resection for ductal adenocarcinoma of the pancreas: 40 years of experience.
      Therefore, an accurate pathologic diagnosis is essential for management of lesions suspicious for pancreatic cancer.
      EUS-guided FNA (EUS-FNA) is a widely used diagnostic tool for pancreatic neoplasms.
      • Haba S.
      • Yamao K.
      • Bhatia V.
      • et al.
      Diagnostic ability and factors affecting accuracy of endoscopic ultrasound-guided fine needle aspiration for pancreatic solid lesions: Japanese large single center experience.
      The diagnostic efficacy of EUS-FNA for solid pancreatic neoplasms has been reported to be good, with a diagnostic accuracy of around 90%; however, there is still some room for improvement.
      • Gonzalo-Marin J.
      • Vila J.J.
      • Perez-Miranda M.
      Role of endoscopic ultrasound in the diagnosis of pancreatic cancer.
      Although rapid on-site evaluation (ROSE) may increase the diagnostic yield,
      • Koul A.
      • Baxi A.C.
      • Shang R.
      • et al.
      The efficacy of rapid on-site evaluation during endoscopic ultrasound-guided fine needle aspiration of pancreatic masses.
      it is not feasible in many institutions because of insufficient medical resources. Therefore, it is necessary to increase the accuracy of diagnosis without ROSE.
      A conventional smear (CS) with Papanicolaou and/or Diff-Quik stain is the most common method for EUS-FNA cytology preparation. However, common problems with CS preparations include bloody smears, dry artifacts, crushing artifacts, and thick tissue fragments, which can obscure the cytologic features and result in a suboptimal diagnosis.
      • Yamabe A.
      • Irisawa A.
      • Bhutani M.
      • et al.
      Efforts to improve the diagnostic accuracy of endoscopic ultrasound-guided fine-needle aspiration for pancreatic tumors.
      Liquid-based cytology (LBC) was developed to overcome the disadvantages associated with CS
      • Linder J.
      Recent advances in thin-layer cytology.
      ; it is currently widely used for cervicovaginal specimens and is increasingly being used for body fluid, urine and respiratory cytology, and FNA cytology from the thyroid and breast.
      • Siebers A.G.
      • Klinkhamer P.J.
      • Grefte J.M.
      • et al.
      Comparison of liquid-based cytology with conventional cytology for detection of cervical cancer precursors: a randomized controlled trial.
      • Wright P.K.
      • Marshall J.
      • Desai M.
      Comparison of SurePath® and ThinPrep® liquid-based cervical cytology using positive predictive value, atypical predictive value and total predictive value as performance indicators.
      • Jeong H.
      • Hong S.R.
      • Chae S.-W.
      • et al.
      Comparison of unsatisfactory samples from conventional smear versus liquid-based cytology in uterine cervical cancer screening test.
      • Hoda R.S.
      • VandenBussche C.
      • Hoda S.A.
      Liquid-based specimen collection, preparation, and morphology. Diagnostic liquid-based cytology.
      • Nishimura R.
      • Aogi K.
      • Yamamoto T.
      • et al.
      Usefulness of liquid-based cytology in hormone receptor analysis of breast cancer specimens.
      • Rossi E.D.
      • Larghi A.
      • Verna E.C.
      • et al.
      Endoscopic ultrasound-guided fine-needle aspiration with liquid-based cytologic preparation in the diagnosis of primary pancreatic lymphoma.
      • Rossi E.D.
      • Zannoni G.F.
      • Moncelsi S.
      • et al.
      Application of liquid-based cytology to fine-needle aspiration biopsies of the thyroid gland.
      However, there are still limited data regarding the diagnostic efficacy of LBC in pancreatic EUS-FNA cytology,
      • LeBlanc J.K.
      • Emerson R.E.
      • Dewitt J.
      • et al.
      A prospective study comparing rapid assessment of smears and ThinPrep® for endoscopic ultrasound-guided fine-needle aspirates.
      • Itonaga M.
      • Murata S.I.
      • Hatamaru K.
      • et al.
      Diagnostic efficacy of smear plus liquid-based cytology for EUS-FNA of solid pancreatic lesions: A propensity-matched study.
      • Hashimoto S.
      • Taguchi H.
      • Higashi M.
      • et al.
      Diagnostic efficacy of liquid-based cytology for solid pancreatic lesion samples obtained with endoscopic ultrasound-guided fine-needle aspiration: propensity score-matched analysis.
      • Qin S.-Y.
      • Zhou Y.
      • Li P.
      • et al.
      Diagnostic efficacy of cell block immunohistochemistry, smear cytology, and liquid-based cytology in endoscopic ultrasound-guided fine-needle aspiration of pancreatic lesions: a single-institution experience.
      • Lee J.K.
      • Choi E.R.
      • Jang T.H.
      • et al.
      A prospective comparison of liquid-based cytology and traditional smear cytology in pancreatic endoscopic ultrasound-guided fine needle aspiration.
      and most studies have used the ThinPrep (Hologic Inc, Marlborough, Mass, USA) method for comparison with CSs.
      • LeBlanc J.K.
      • Emerson R.E.
      • Dewitt J.
      • et al.
      A prospective study comparing rapid assessment of smears and ThinPrep® for endoscopic ultrasound-guided fine-needle aspirates.
      ,
      • Itonaga M.
      • Murata S.I.
      • Hatamaru K.
      • et al.
      Diagnostic efficacy of smear plus liquid-based cytology for EUS-FNA of solid pancreatic lesions: A propensity-matched study.
      ,
      • Qin S.-Y.
      • Zhou Y.
      • Li P.
      • et al.
      Diagnostic efficacy of cell block immunohistochemistry, smear cytology, and liquid-based cytology in endoscopic ultrasound-guided fine-needle aspiration of pancreatic lesions: a single-institution experience.
      ,
      • Lee J.K.
      • Choi E.R.
      • Jang T.H.
      • et al.
      A prospective comparison of liquid-based cytology and traditional smear cytology in pancreatic endoscopic ultrasound-guided fine needle aspiration.
      Unsatisfactory rates have been reported to be lower by the SurePath (BD Diagnostics, Burlington, NC, USA) method compared with ThinPrep for cervicovaginal cytology specimens,
      • Fontaine D.
      • Narine N.
      • Naugler C.
      Unsatisfactory rates vary between cervical cytology samples prepared using ThinPrep and SurePath platforms: a review and meta-analysis.
      therefore we sought to evaluate the diagnostic performance of SurePath LBC for diagnosing pancreatic cancers in EUS-FNA specimens in comparison with CSs.

      Methods

      Study design and participants

      To evaluate the diagnostic performance of LBC, a randomized, prospective, comparative, single-center study was conducted at Seoul National University Hospital between April 2018 and March 2019 (Fig. 1). Patients >18 years of age who underwent EUS-FNA for suspected pancreatic malignancy based on imaging findings were included in this study. Patients were excluded from the study if they did not give consent or had serious mental illness. In addition, patients with severe cardiopulmonary diseases or coagulopathy and pregnant patients were excluded because unexpected events during the procedure could result in severe adverse events in these high-risk patients. Patients were also excluded if the target lesion was difficult to approach by EUS as a result of previous thoracoabdominal surgery or if there was a risk of tract seeding after the examination.
      • Yamao K.
      • Sawaki A.
      • Mizuno N.
      • et al.
      Endoscopic ultrasound-guided fine-needle aspiration biopsy (EUS-FNAB): past, present, and future.
      ,
      • Erickson R.A.
      EUS-guided FNA.
      Figure thumbnail gr1
      Figure 1Flowchart of patients throughout the randomized trial. CS, Conventional smear; LBC, liquid-based cytology.
      The study protocol was approved by the institutional review board of Seoul National University Hospital and was conducted in accordance with the provisions of the Declaration of Helsinki. All participants provided written informed consent before recruitment. The trial is registered with ClinicalTrials.gov as NCT03606148.

      EUS-FNA procedure

      A linear EUS scope (GF-UCT260; Olympus Medical Systems, Tokyo, Japan) with a 19- or 22-gauge needle (EZ Shot 3 Plus; Olympus Medical Systems, Tokyo, Japan) was used at the discretion of the endoscopist. All EUS-FNA procedures were done by 4 experienced endoscopists, who have performed more than 500 EUS-FNA procedures. When the target lesion in the pancreas was visualized by EUS, the needle was punctured into the lesion using an elevator at the discretion of endoscopist. After the stylet was removed, the needle was moved forward and backward 15 times with an attached 10-mL suction syringe. The tissue sample was collected through 3 needle passes. If tissue core biopsy specimens were determined to be insufficient on gross inspection, additional needle passes for core biopsy were performed at the discretion of the endoscopist.
      Enrolled patients were randomly assigned on a 1:1 basis to the LBC group or the CS group before the EUS-FNA procedure by personnel not involved in the trial. Block randomization with randomly selected block sizes generated by Microsoft Excel 2013 (Microsoft Corp., Redmond, Wash, USA) was adopted. In the LBC group, specimens obtained by the first needle pass were subjected to LBC and specimens from the second needle pass were used for CS. The aspirates were prepared the other way around for the CS group, ie, CS slides were prepared first, followed by LBC. For both groups, tissue core biopsy specimens for histologic evaluation were collected from the third and subsequent needle passes.

      Specimen acquisition and evaluation

      For LBC, the specimens were immediately suspended in preservative fluid (CytoRich Red, Thermo Scientific, Waltham, Mass, USA) and sent to the pathology department. Specimens were further processed according to the SurePath NON-GYN protocol on the BD PrepStain Slide Processor (Becton, Dickinson abd Company, Franklin Lakes, NJ, USA).
      • Hoda R.S.
      • VandenBussche C.
      • Hoda S.A.
      Liquid-based specimen collection, preparation, and morphology. Diagnostic liquid-based cytology.
      The slides were stained with Papanicolaou stain. Aspirated specimens from the CS group were smeared immediately onto glass slides and fixed in 95% alcohol by an experienced endoscopy nurse who has performed slide preparations for more than 100 cases annually. Core biopsy specimens were prepared by dropping the specimens onto a slide and transferring them into 10% formalin if visible tissue was present. The pathologists were blinded to each other’s assessments, and the parameters were initially recorded independently in 2 separate electronic files. The data were subsequently merged into 1 file by 1 pathologist (H.K.), and all discrepant cases were reviewed and discussed by both pathologists using a multiheaded microscope to render a final assessment.
      The EUS-FNA cytology results were classified as inadequate, benign, atypical, suspicious, or malignant. For statistical purposes, we classified suspicious and malignant as malignant, and atypical cases as benign, as previously described.
      • LeBlanc J.K.
      • Emerson R.E.
      • Dewitt J.
      • et al.
      A prospective study comparing rapid assessment of smears and ThinPrep® for endoscopic ultrasound-guided fine-needle aspirates.
      ,
      • Itonaga M.
      • Murata S.I.
      • Hatamaru K.
      • et al.
      Diagnostic efficacy of smear plus liquid-based cytology for EUS-FNA of solid pancreatic lesions: A propensity-matched study.
      ,
      • Yeon M.H.
      • Jeong H.S.
      • Lee H.S.
      • et al.
      Comparison of liquid-based cytology (CellPrepPlus) and conventional smears in pancreaticobiliary disease.
      We examined the following parameters for the individual LBC and CS cases: (1) specimen adequacy (adequate, material from target organ present and sufficient to provide a cytologic diagnosis; inadequate, insufficient material for diagnosis); (2) presence of a dry artifact; (3) the background of the slide preparation (clean, bloody, inflammatory, mucinous, necrotic); (4) presence of blood clots on the slide; (5) presence of mucin on the slide; (6) presence of contaminants (GI, squamous, mesothelial); (7) tumor cellularity (0, no cells; 1, <3 clusters; 2, 3-10 clusters; 3, 10-20 clusters; 4, >20 clusters); (8) the architectural pattern (2-dimensional monolayer sheets, 3-dimensional clusters, predominantly scattered individual cells); (9) nuclear pleomorphism; (10) conspicuous nucleoli; (11) hyperchromasia; and (12) coarse chromatin.
      The final diagnosis was based on LBC, CS, EUS-FNA core biopsy, or, if available, the pathologic diagnosis of the surgically resected specimens. If any of these specimens were evaluated as malignant or suspicious, the final diagnosis was reported as malignant, and if all specimens were assessed as atypical or benign, the lesion was diagnosed as benign. If the initial diagnosis was benign, a final diagnosis was made based on clinical judgement after a follow-up period of at least 6 months with necessary radiologic studies.

      Study outcomes

      As the primary outcome, we compared the diagnostic performance (accuracy, sensitivity, specificity, positive predictive value, and negative predictive value) between CS and LBC based on the final diagnosis as the criterion standard. In addition, we evaluated the cytomorphologic features of each method as secondary outcomes.

      Statistical analysis

      We adopted the noninferiority principle to calculate the sample size in this study. A previous study comparing LBC with CS estimated the diagnostic accuracy of LBC and CS to be approximately 92% and 90%, respectively.
      • Hafez N.H.
      • Shaaban H.M.
      SurePath liquid-based cytology versus conventional smears for interpretation of serous effusion fluids: a study of 104 cases.
      When the noninferiority margin (−Δ) was considered as −10%, we estimated that 170 patients would provide 80% power with a 2-sided alpha level of 0.05, considering a drop-out rate of 15%. Noninferiority would be declared if the lower limit of the 95% confidence interval of the proportion difference between the diagnostic accuracy of CS and LBC was not below −10% of the noninferior margin.
      The difference between LBC and CS was evaluated by means of the Student t test for continuous variables and the chi-squared test and Fisher exact test for categorical variables. Values of P < .05 were considered statistically significant and all P values were 2-sided. All statistical analyses were conducted with SPSS software version 18.0 (SPSS Inc, Chicago, Ill, USA).

      Results

      Baseline characteristics of the study population

      A total of 170 patients with suspected pancreatic malignancy were enrolled. The mean age was 64.8 years (range, 37-88 years; standard deviation, 10.6 years), and 95 (55.9%) were male. The median size of the target lesions was 3.0 cm (range, 1.3-11.0 cm), and the masses were located in the head, uncinate process, body, and tail in 59 (34.7%), 24 (14.1%), 38 (22.4%), and 40 (23.5%) cases, respectively. During the EUS-FNA procedure, tissue samples were acquired with 19-gauge or 22-gauge needles, mostly using 3 needle passes. The baseline clinical characteristics were similar in both groups (Table 1).
      Table 1Baseline characteristics of the study population
      VariableTotalConventional smear groupLiquid-based cytology groupP value
      Number of patients1708585
      Sex, male/female95/7547/3848/37.877
      Age (years), mean ± SD (range)64.8 ± 10.6 (37-88)64.5 ± 10.865.3 ± 10.5.619
      Diagnosis confirmation
       At first attempt, n (%)156 (91.8)77 (90.6)79 (92.9).577
       At second attempt, n (%)165 (97.1)81 (95.3)84 (98.8).368
      Needle gauge, 19/22 gauge51/11932/5319/66.03
      Number of needle passes, 3/4146/1977/769/12.192
      Size (cm), mean ± SD (range)3.2 ± 1.1 (1.3-11.0)3.1 ± 1.23.3 ± 1.1.286
      Location, n (%).243
       Head59 (34.7)3029
       Uncinate24 (14.1)1212
       Body43 (25.3)2617
       Tail44 (25.9)1727
      Adverse events, n (%)13 (7.6)49.149
       Abdominal pain4 (2.4)22
       Bleeding5 (2.9)23
       Fever3 (1.8)03
       Perforation1 (0.6)01
      Final diagnosis, n (%)
       Benign5 (2.9)32.999
       Malignant165 (97.1)8283.443
       Pancreatic ductal adenocarcinoma158 (92.9)8078
       IPMN with invasive carcinoma3 (1.8)03
       MCN with invasive carcinoma1 (0.6)10
       Neuroendocrine tumor2 (1.2)02
       Metastatic carcinoma in lymph node1 (0.6)10
      SD, Standard deviation; IPMN, intraductal papillary mucinous neoplasm; MCN, mucinous cystic neoplasm.
      Of the 170 cases, 165 (97.1%) were classified as malignant in the final diagnosis (Table 1); 156 cases (91.8%) were diagnosed at the first EUS-FNA attempt, and 9 cases were diagnosed after a second EUS-FNA attempt. Of the remaining 5 cases in which a diagnosis of malignancy was not rendered by EUS-FNA cytology, 3 were finally classified as malignant on the surgically resected specimens, and 2 were diagnosed after 6-month imaging and clinical follow-up. Of the 165 malignant cases by EUS-FNA, 1 was a metastatic lymph node (metastatic acinar cell carcinoma from the pancreas) and was therefore excluded from further analysis.

      Diagnostic performance of conventional smear and liquid-based cytology

      The cytologic diagnoses of the CS and LBC cases and the final diagnoses are summarized in Table 2. The diagnostic accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of LBC versus CS were 88.0% versus 83.8%, 87.7% versus 83.2%, 100% versus 100%, 100% versus 100%, and 16.7% versus 16.1%, respectively (Table 3). The difference in diagnostic accuracy between CS and LBC was 4.2%, and the lower limit of the confidence interval was more than −10% of the noninferiority margin (Fig. 2). These results suggest that the diagnostic efficacy of LBC is not inferior to that of CS. In addition, when the LBC results were combined with the core biopsy results, the diagnostic accuracy and sensitivity were significantly improved compared with that of LBC alone (95.3% vs 88.0% and 95.1% vs 87.7%; P = .016 and P = .017, respectively). Inadequate specimens for diagnosis were less frequent for LBC compared with CS (3 of 170 vs 9 of 170; P = .078).
      Table 2Comparison of final diagnoses and results of the cytologic tests
      DiagnosisFinal diagnosis with conventional smear (n)Final diagnosis with liquid-based cytology (n)
      BenignMalignantTotalBenignMalignantTotal
      Malignant01071070117117
      Suspicious0222202525
      Atypical2171911415
      Benign3912369
      Inadequate099123
      Total51641695164169
      Table 3Diagnostic efficacy of conventional smear cytology versus liquid-based cytology
      Conventional smear (95% CI)Liquid-based cytology (95% CI)Difference (95% CI)P value
      Accuracy, %83.8 (80.5-83.8)88.0 (85.1-88.0)4.2 (−3.4 to 11.9).276
      Sensitivity, %83.2 (81.5-83.2)87.7 (86.2-87.7)4.5 (−3.3 to 12.4).256
      Specificity, %100 (47.4-100)100 (40.5-100)0 (−43.4 to 49.0).999
      Positive predictive value, %100 (98.3-100)100 (98.3-100)0 (−2.9 to 2.6).999
      Negative predictive value, %16.1 (7.6-16.1)16.7 (6.8-16.7)0.6 (-18.7 to 21.8).953
      CI, Confidence interval.
      Figure thumbnail gr2
      Figure 2Analyzing noninferiority of liquid-based cytology to conventional smears.

      Cytomorphologic features of conventional smear and liquid-based cytology

      CS preparations resulted in an average of 5.23 slides per case (range, 1-16 slides), whereas all LBC cases resulted in 1 slide (P < .001). The aspirates were concentrated within a 1.3-mm diameter circle on the LBC slides. The cytomorphologic features of the CS and LBC cases are summarized in Table 4. Air-dry artifacts were noted only in the CS cases (CS, 4.1%; LBC, 0%; P = .015). A significantly larger number of LBC cases (71.0%) showed a clean background compared with CS (5.3%; P < .001). A bloody background was significantly more frequent in CS preparations (CS, 85.2%; LBC, 1.8%; P < .001) and clotted blood was also frequently seen in CS cases (63.9%), obscuring the cytologic features and allowing the introduction of air bubbles beneath the cover slides due to the irregular thickness of the smear (Fig. 3A and B). Mucin was significantly reduced on the LBC slides (7.1%) compared with CS (20.1%; P < .001) (Fig. 3C and D). There was no significant difference in the frequency of an inflammatory or necrotic background between LBC and CS. Contaminants were more frequently seen in CS preparations (68.0%) compared with LBC (50.3%; P = .001). There were no significant differences between LBC and CS with regard to the cytologic characteristics of the tumor cells. The cytomorphologic features were similar between the 2 groups: the architecture of the tumor cell groups (2-dimensional monolayer sheets, 3-dimensional clusters, or individually scattered cells) were similar, and the nuclear features characteristic of malignant cells were present in both groups (Fig. 4). However, as a whole, the tumors cells on the LBC slides demonstrated more prominent nucleoli compared with CS preparations.
      Table 4Comparison of cytomorphologic features between conventional smear and liquid-based cytology
      Cytomorphologic featuresConventional smear, n/N (%)Liquid-based cytology, n/N (%)P value
      Quality and background
      Dry artifact7/169 (4.1)0/169 (0).015
      Background
       Clean9/169 (5.3)120/169 (71.0)<.001
       Bloody144/169 (85.2)3/169 (1.8)<.001
       Mucinous34/169 (20.1)12/169 (7.1)<.001
       Inflammatory11/169 (6.5)19/169 (11.2).126
       Necrotic31/169 (18.3)19/169 (11.2).066
      Blood clots present108/169 (63.9)1/169 (0.6)<.001
      Mucin present93/169 (55.0)69/169 (40.8).009
      Contaminants present115/169 (68.0)85/169 (50.3).001
       Gastrointestinal112/169 (66.3)74/169 (43.8)<.001
       Intestinal10/169 (5.9)12/169 (7.1).659
       Squamous3/169 (1.8)10/169 (5.9).048
       Mesothelial0/169 (0)0/169 (0).999
      Tumor cell characteristics
      Neuroendocrine tumors (n = 2) were included in this analysis.
      Cellularity
       Acellular10/169 (5.9)3/169 (1.8).048
       Sparsely cellular34/169 (20.1)37/169 (21.9).689
       Moderately cellular26/169 (15.4)23/169 (13.6).643
       Highly cellular49/169 (29.0)43/169 (25.4).463
       Very highly cellular50/169 (29.6)63/169 (37.3).134
      Monolayer sheets present92/169 (54.4)81/169 (47.9).231
      Three-dimensional clusters present142/169 (84.0)145/169 (85.8).648
      Single-cell predominance20/169 (11.8)29/169 (17.2).164
      Nuclear pleomorphism122/151 (80.8)131/158 (82.9).629
      Conspicuous nucleoli118/150 (78.7)134/158 (84.8).162
      Hyperchromasia135/150 (90.0)140/158 (88.6).693
      Coarse chromatin134/150 (89.3)140/158 (88.6).839
      Neuroendocrine tumors (n = 2) were included in this analysis.
      Figure thumbnail gr3
      Figure 3Cytologic features of conventional smears and liquid-based cytology at low-power magnification (A-D, orig. mag., ×40). Conventional smears were frequently contaminated by blood, obscuring evaluation of the cellular elements (A), whereas blood was reduced in the liquid-based cytology slides, resulting in clean backgrounds (B). On the other hand, thick mucin was retained in conventional smears (C, a case of mucinous carcinoma), whereas the mucin was more watery and scant in the liquid-based preparation (D).
      Figure thumbnail gr4
      Figure 4Cytologic features of conventional smears and liquid-based cytology at high-power magnification (A-C, orig. mag., ×400). The cytologic features were similar between conventional smears (A) and liquid-based cytology (B). The nuclear features of malignancy (nuclear pleomorphism, hyperchromasia, and prominent nucleoli) and the clear cytoplasm in this case of ductal adenocarcinoma are present in both slides. A liquid-based preparation of another example of a well-differentiated ductal adenocarcinoma (C) demonstrates mild nuclear pleomorphism and prominent red nucleoli.
      Two cases were diagnosed as pancreatic neuroendocrine tumor (PanNET). Both LBC and CS slides demonstrated neoplastic cells with the typical morphologic features of PanNET, including loose aggregates of tumor cells with individually scattered cells in the background, relatively uniform nuclei, nuclear hyperchromasia, and coarse salt-and-pepper chromatin. The cytomorphology of PanNET was similar in the CS and LBC preparations for both cases. The cytomorphologic features of the LBC and CS slides were similar regardless of the order of specimen acquisition (ie, allocation to the LBC and CS groups).

      Discussion

      In this prospective randomized crossover trial, we evaluated the diagnostic value of LBC compared with CS for the cytologic diagnosis of solid pancreatic masses, and we demonstrate that the diagnostic efficacy of LBC is not inferior to that of CS. LBC resulted in a lower frequency of inadequate specimens and significantly reduced blood contamination, whereas the cytologic features of malignancy were preserved. In addition, the accuracy and sensitivity of LBC were significantly improved when combined with the core biopsy results. Because an accurate diagnosis of pancreatic cancer on EUS-FNA cytology is of critical importance in the management of patients with this lethal disease, these results could be encouraging for both gastroenterologists and cytopathologists, especially in institutions where ROSE is not feasible.
      LBC is popular for the following reasons: (1) reduction in dry artifacts and better preservation of the specimens due to immediate fixation in the LBC medium; (2) elimination of blood contamination by lysing red blood cells and resulting in a cleaner background; and (3) an even distribution of the cellular elements within a circle in the center of the slide, thus less time-consuming for the cytopathologist compared with CS where the samples are irregularly distributed over multiple glass slides.
      • Hoda R.S.
      • VandenBussche C.
      • Hoda S.A.
      Liquid-based specimen collection, preparation, and morphology. Diagnostic liquid-based cytology.
      ,
      • Yeon M.H.
      • Jeong H.S.
      • Lee H.S.
      • et al.
      Comparison of liquid-based cytology (CellPrepPlus) and conventional smears in pancreaticobiliary disease.
      However, despite these known advantages, LBC is not commonly performed for pancreatic EUS-FNA specimens, and the diagnostic efficacy of LBC compared with CS in the literature is still controversial. Some previous prospective studies reported that the CS method was more sensitive and accurate than LBC for diagnosing pancreatic malignancies,
      • LeBlanc J.K.
      • Emerson R.E.
      • Dewitt J.
      • et al.
      A prospective study comparing rapid assessment of smears and ThinPrep® for endoscopic ultrasound-guided fine-needle aspirates.
      ,
      • Lee J.K.
      • Choi E.R.
      • Jang T.H.
      • et al.
      A prospective comparison of liquid-based cytology and traditional smear cytology in pancreatic endoscopic ultrasound-guided fine needle aspiration.
      ,
      • Yeon M.H.
      • Jeong H.S.
      • Lee H.S.
      • et al.
      Comparison of liquid-based cytology (CellPrepPlus) and conventional smears in pancreaticobiliary disease.
      whereas 1 retrospective study demonstrated superior diagnostic accuracy and sensitivity for LBC compared with CS.
      • Hashimoto S.
      • Taguchi H.
      • Higashi M.
      • et al.
      Diagnostic efficacy of liquid-based cytology for solid pancreatic lesion samples obtained with endoscopic ultrasound-guided fine-needle aspiration: propensity score-matched analysis.
      Interestingly, the latter study used the SurePath system, whereas the other studies used the ThinPrep method. It has been demonstrated that the SurePath method resulted in fewer unsatisfactory specimens compared with ThinPrep in various studies on cervicovaginal specimens; however, direct comparisons between ThinPrep and SurePath have not been performed for pancreatic EUS-FNA specimens.
      • Fontaine D.
      • Narine N.
      • Naugler C.
      Unsatisfactory rates vary between cervical cytology samples prepared using ThinPrep and SurePath platforms: a review and meta-analysis.
      ,
      • Bigras G.
      • Rieder M.A.
      • Lambercy J.M.
      • et al.
      Keeping collecting device in liquid medium is mandatory to ensure optimized liquid-based cervical cytologic sampling.
      • Naeem R.C.
      • Goldstein D.Y.
      • Einstein M.H.
      • et al.
      SurePath specimens versus ThinPrep specimen types on the COBAS 4800 platform: high-risk HPV status and cytology correlation in an ethnically diverse Bronx population.
      • Rozemeijer K.
      • Penning C.
      • Siebers A.G.
      • et al.
      Comparing SurePath, ThinPrep, and conventional cytology as primary test method: SurePath is associated with increased CIN II+ detection rates.
      Although a reduction of background material by fixation media (eg, ethanol-based CytoRich Red for SurePath, methanol-based CytoLyt solution for ThinPrep) is a common principle for LBC, the methods of processing the specimens are different. ThinPrep uses a vacuum filtration technique, whereas SurePath uses a sedimentation process.
      • Hoda R.S.
      • VandenBussche C.
      • Hoda S.A.
      Liquid-based specimen collection, preparation, and morphology. Diagnostic liquid-based cytology.
      Before the sedimentation process, the original SurePath method for cervicovaginal cytology uses a density gradient reagent that substantially reduces the background inflammatory cells in addition to the blood, whereas CytoRich Red solution is used for nongynecologic specimens to preserve the cellular components and eliminate the blood.
      In accordance with previous studies, our data demonstrate a reduced number of inadequate specimens and a significantly higher number of preparations with clean backgrounds in the LBC group compared with the CS group.
      • Itonaga M.
      • Murata S.I.
      • Hatamaru K.
      • et al.
      Diagnostic efficacy of smear plus liquid-based cytology for EUS-FNA of solid pancreatic lesions: A propensity-matched study.
      ,
      • Hashimoto S.
      • Taguchi H.
      • Higashi M.
      • et al.
      Diagnostic efficacy of liquid-based cytology for solid pancreatic lesion samples obtained with endoscopic ultrasound-guided fine-needle aspiration: propensity score-matched analysis.
      ,
      • Yeon M.H.
      • Jeong H.S.
      • Lee H.S.
      • et al.
      Comparison of liquid-based cytology (CellPrepPlus) and conventional smears in pancreaticobiliary disease.
      Even though the background was reduced, the diagnostically relevant cells were preserved in the LBC slides, resulting in more efficient and easier cytologic evaluation. In addition, the cytomorphologic features of malignancy, such as nuclear pleomorphism, prominent nucleoli, and nuclear hyperchromasia, were preserved in the LBC specimens. Even in well-differentiated ductal adenocarcinomas where the nuclear changes are more subtle, mild nuclear pleomorphism and loss of polarity could still be appreciated in the LBC slides, and prominent red nucleoli were also frequently seen. However, for adenocarcinomas with abundant mucin production, the thick mucin present on the CS preparations was more diluted and scant in the LBC slides. Therefore, although LBC is a promising method for cytologic evaluation of pancreatic solid masses, it may have limited use for pancreatic cystic lesions where the background extracellular material may be reduced by the LBC processing methods.
      In addition to the aforementioned advantages of LBC from the viewpoint of diagnostic performance, this method is also easy to prepare in the endoscopy suite, resulting in decreased procedure time for the gastroenterologists. For CSs, the aspirated sample is smeared directly onto several glass slides by the endoscopy nurse in the EUS examination room, and obtaining good quality smears requires a lot of practice and dexterity. On the other hand, for LBC, the aspirated material is simply dropped into a collection vial containing the fixative solution provided by the manufacturer, which is sent to the cytopathology laboratory for automated processing.
      • Hoda R.S.
      • VandenBussche C.
      • Hoda S.A.
      Liquid-based specimen collection, preparation, and morphology. Diagnostic liquid-based cytology.
      ,
      • Sweeney B.J.
      • Haq Z.
      • Happel J.F.
      • et al.
      Comparison of the effectiveness of two liquid-based Papanicolaou systems in the handling of adverse limiting factors, such as excessive blood.
      Another advantage of LBC is that cell blocks can be prepared from residual samples, and these cell blocks may be of additional value in the diagnosis of pancreatic cancer by providing an extra biopsy sample stained with hematoxylin and eosin for further histologic evaluation and by allowing for ancillary immunocytochemical stains (eg, neuroendocrine markers). In our practice, cell blocks are routinely prepared from all LBC samples in addition to SurePath slides, which helps to maximize the diagnostic yield of the EUS-FNA specimens. Furthermore, in addition to cell block preparation, the residual LBC specimen can be subjected to molecular analysis when necessary.
      • Linder J.
      Recent advances in thin-layer cytology.
      ,
      • Hoda R.S.
      • VandenBussche C.
      • Hoda S.A.
      Liquid-based specimen collection, preparation, and morphology. Diagnostic liquid-based cytology.
      There are some limitations in this study. The protocol for the EUS-FNA procedure was not completely uniform. The needle gauge used for sampling varied depending on the operator and the nature of the target lesions. However, although a certain gauge needle was more frequently used, the same needle was used for both preparation methods for each case in this crossover trial. Therefore, when analyzing the differences in the diagnostic performance between the 2 preparation methods, the difference in the type of needle used would be negligible. In addition, we found no differences in the diagnostic performance (Supplementary Table 1, available online at www.giejournal.org) or the cytomorphologic features (Supplementary Table 2, available online at www.giejournal.org) in a subgroup analysis according to the type of needle used.
      Because the endoscopists in this study perform more than 100 EUS-FNA procedure annually, the skill of the endoscopists made no significant difference in the specimen quality. With regard to the procedure for smear preparation, EUS-FNA smear slides in our institution are prepared by an experienced endoscopy nurse who has performed more than 100 cases annually under the endoscopists’ supervision. A previous study on the diagnostic performance of CSs prepared by the endoscopy nurse at this institution demonstrated a sensitivity of 87.9% and accuracy of 88.1%, which are similar to the present study and several previous studies.
      • Itonaga M.
      • Murata S.I.
      • Hatamaru K.
      • et al.
      Diagnostic efficacy of smear plus liquid-based cytology for EUS-FNA of solid pancreatic lesions: A propensity-matched study.
      ,
      • Qin S.-Y.
      • Zhou Y.
      • Li P.
      • et al.
      Diagnostic efficacy of cell block immunohistochemistry, smear cytology, and liquid-based cytology in endoscopic ultrasound-guided fine-needle aspiration of pancreatic lesions: a single-institution experience.
      ,
      • Jang D.K.
      • Lee S.H.
      • Lee J.K.
      • et al.
      Comparison of cytological and histological preparations in the diagnosis of pancreatic malignancies using endoscopic ultrasound-guided fine needle aspiration.
      Finally, the cost difference between the 2 methods was not assessed in this study. Although the LBC method is more expensive than CS, the reduction in inadequate smears alone would make LBC less expensive in the long run compared with CS.
      • Denton K.J.
      Liquid based cytology in cervical cancer screening.
      Further studies regarding the cost-effectiveness of LBC would be helpful to determine the applicability of LBC in resource-limited settings.
      Nevertheless, the strength of our study is that this is a prospective randomized crossover study with a large sample size, performed at a large-volume tertiary medical institution with experienced endoscopists and pancreaticobiliary expert pathologists. As far as we know, this is the first prospective randomized study using SurePath to compare the diagnostic yield between LBC and CS in pancreatic EUS-FNA, in addition to a detailed cytomorphologic comparison between the 2 methods.
      In conclusion, the diagnostic performance of LBC was not inferior to CS for the diagnosis of pancreatic malignancies on EUS-FNA specimens. The reduction in bloody backgrounds facilitated the identification of diagnostically relevant cells on LBC slides, and the cytomorphologic features of the tumors were maintained on LBC slides, justifying the use of LBC for EUS-FNA over CS when available.

      Acknowledgments

      The authors are grateful to Hyoung Jin An for technical assistance with the preparation of the cytology slides.

      Appendix

      Supplementary Table 1Diagnostic efficacy of conventional smear cytology versus liquid-based cytology according to the needle gauge
      19-gauge needle22-gauge needle
      LBCCSLBCCS
      There was also no significant difference in diagnostic efficacy between LBC using a 19-gauge needle and CS with a 22-guage needle (P = .157).
      Accuracy, %90.089.887.181.1
      Sensitivity, %89.689.486.880.6
      Specificity, %100100100100
      Positive predictive value, %100100100100
      Negative predictive value, %28.628.611.812.5
      LBC, Liquid-based cytology; CS, conventional smear.
      There was also no significant difference in diagnostic efficacy between LBC using a 19-gauge needle and CS with a 22-guage needle (P = .157).
      Supplementary Table 2Comparison of cytomorphologic features according to the preparation methods and needle gauge
      Cytomorphologic featuresLiquid-based cytologyConventional smear
      Total, n/N19G needle, n/N (%)22G needle, n/N (%)P valueTotal, n/N19G needle, n/N (%)22G needle, n/N (%)P value
      Quality and background
      Inadequate3/1690/50 (0)3/119 (2.5).55619/1694/50 (8.0)15/119 (12.6).387
      Dry artifact0/1690/50 (0)0/119 (0)7/1692/50 (4.0)5/119 (4.2).999
      Background
       Clean120/16932 (64)88 (73.9).1939/1691/50 (2.0)8/119 (6.7).284
       Bloody3/1690 (0)3 (2.5).556144/16941/50 (82.0)103/119 (86.6).447
       Inflammatory19/1699 (18)10 (8.4).07111/1695/50 (10.0)6/119 (5.0).305
       Mucinous12/1694 (8.0)8 (6.7).7534/16914/50 (28.0)20/119 (16.8).098
       Necrotic19/1696 (12)13 (10.9).8431/16912/50 (24.0)19/119 (16.0).218
      Blood clots present1/1690 (0)1 (0.8).999108/16932/50 (64.0)76/119 (63.9).987
      Mucin present69/16924 (48)45 (37.8).21993/16932/50 (64.0)61/119 (51.3).129
      Contaminants present85/1622 (44)63 (52.9).289115/16939/50 (78.0)76/119 (63.9).072
       Gastrointestinal74/16917 (34)57 (47.9).096112/16937/50 (74.0)75/119 (63.0).168
       Intestinal12/1695 (10)7 (5.9).34110/1697/50 (14.0)3/119 (2.5).008
       Squamous10/1696 (12)4 (3.4).0663/1692/50 (4.0)1/119 (0.8).209
       Mesothelial0/1690 (0)0 (0)0/1690/50 (0)0/119 (0)
      Tumor cell characteristics
      Cellularity
       Acellular3/1690 (0)3 (2.5).55610/1692/50 (4.0)8/119 (6.7).725
       Sparsely cellular37/1699 (18)28 (23.5).42834/1697/50 (14.0)27/119 (22.7).198
       Moderately cellular23/1696 (12)17 (14.3).69226/1698/50 (16.0)18/119 (15.1).886
       Highly cellular43/16912 (24)31 (26.1).7849/16914/50 (28.0)35/119 (29.4).854
       Very highly cellular63/16923/50 (46)40/119 (33.6).12950/16919/50 (38.0)31/119 (26.1).12
      Monolayer sheets present81/16927/50 (54)54/119 (45.4).30692/16929/50 (58.0)63/119 (52.9).547
      Three-dimensional clusters present145/16944/50 (88)101/119 (84.9).595142/16945/50 (90.0)97/119 (81.5).169
      Single-cell predominance29/1697/50 (14)22/119 (18.5).4820/1696/50 (12.0)14/119 (11.8).966
      Nuclear pleomorphism131/15839/45 (86.7)92/113 (81.4).429122/15138/46 (82.6)84/105 (80.0).708
      Conspicuous nucleoli134/15841/45 (91.1)93/113 (82.3).164118/15037/46 (80.4)81/104 (77.9).725
      Hyperchromasia140/15840/45 (88.9)100/113 (88.5).944135/15042/46 (91.3)93/104 (89.4).999
      Coarse chromatin140/15839/45 (86.7)101/113 (89.4).628134/15041/46 (89.1)93/104 (89.4).999

      References

        • Jung K.-W.
        • Won Y.-J.
        • Kong H.-J.
        • et al.
        Cancer statistics in Korea: incidence, mortality, survival, and prevalence in 2015.
        Cancer Res Treat. 2018; 50: 303-316
        • Saad A.M.
        • Turk T.
        • Al-Husseini M.J.
        • et al.
        Trends in pancreatic adenocarcinoma incidence and mortality in the United States in the last four decades; a SEER-based study.
        BMC Cancer. 2018; 18: 688
        • Dusch N.
        • Weiss C.
        • Strobel P.
        • et al.
        Factors predicting long-term survival following pancreatic resection for ductal adenocarcinoma of the pancreas: 40 years of experience.
        J Gastrointest Surg. 2014; 18: 674-681
        • Haba S.
        • Yamao K.
        • Bhatia V.
        • et al.
        Diagnostic ability and factors affecting accuracy of endoscopic ultrasound-guided fine needle aspiration for pancreatic solid lesions: Japanese large single center experience.
        J Gastroenterol. 2013; 48: 973-981
        • Gonzalo-Marin J.
        • Vila J.J.
        • Perez-Miranda M.
        Role of endoscopic ultrasound in the diagnosis of pancreatic cancer.
        World J Gastrointest Oncol. 2014; 6: 360-368
        • Koul A.
        • Baxi A.C.
        • Shang R.
        • et al.
        The efficacy of rapid on-site evaluation during endoscopic ultrasound-guided fine needle aspiration of pancreatic masses.
        Gastroenterol Rep. 2017; 6: 45-48
        • Yamabe A.
        • Irisawa A.
        • Bhutani M.
        • et al.
        Efforts to improve the diagnostic accuracy of endoscopic ultrasound-guided fine-needle aspiration for pancreatic tumors.
        Endosc Ultrasound. 2016; 5: 225-232
        • Linder J.
        Recent advances in thin-layer cytology.
        Diagn Cytopathol. 1998; 18: 24-32
        • Siebers A.G.
        • Klinkhamer P.J.
        • Grefte J.M.
        • et al.
        Comparison of liquid-based cytology with conventional cytology for detection of cervical cancer precursors: a randomized controlled trial.
        JAMA. 2009; 302: 1757-1764
        • Wright P.K.
        • Marshall J.
        • Desai M.
        Comparison of SurePath® and ThinPrep® liquid-based cervical cytology using positive predictive value, atypical predictive value and total predictive value as performance indicators.
        Cytopathology. 2010; 21: 374-378
        • Jeong H.
        • Hong S.R.
        • Chae S.-W.
        • et al.
        Comparison of unsatisfactory samples from conventional smear versus liquid-based cytology in uterine cervical cancer screening test.
        J Pathol Transl Med. 2017; 51: 314-319
        • Hoda R.S.
        • VandenBussche C.
        • Hoda S.A.
        Liquid-based specimen collection, preparation, and morphology. Diagnostic liquid-based cytology.
        Springer, Berlin, Heidelberg2017: 1-12
        • Nishimura R.
        • Aogi K.
        • Yamamoto T.
        • et al.
        Usefulness of liquid-based cytology in hormone receptor analysis of breast cancer specimens.
        Virchows Arch. 2011; 458: 153-158
        • Rossi E.D.
        • Larghi A.
        • Verna E.C.
        • et al.
        Endoscopic ultrasound-guided fine-needle aspiration with liquid-based cytologic preparation in the diagnosis of primary pancreatic lymphoma.
        Pancreas. 2010; 39: 1299-1302
        • Rossi E.D.
        • Zannoni G.F.
        • Moncelsi S.
        • et al.
        Application of liquid-based cytology to fine-needle aspiration biopsies of the thyroid gland.
        Front Endocrinol. 2012; 3: 57
        • LeBlanc J.K.
        • Emerson R.E.
        • Dewitt J.
        • et al.
        A prospective study comparing rapid assessment of smears and ThinPrep® for endoscopic ultrasound-guided fine-needle aspirates.
        Endoscopy. 2010; 42: 389-394
        • Itonaga M.
        • Murata S.I.
        • Hatamaru K.
        • et al.
        Diagnostic efficacy of smear plus liquid-based cytology for EUS-FNA of solid pancreatic lesions: A propensity-matched study.
        Medicine (Baltimore). 2019; 98e15575
        • Hashimoto S.
        • Taguchi H.
        • Higashi M.
        • et al.
        Diagnostic efficacy of liquid-based cytology for solid pancreatic lesion samples obtained with endoscopic ultrasound-guided fine-needle aspiration: propensity score-matched analysis.
        Dig Endosc. 2017; 29: 608-616
        • Qin S.-Y.
        • Zhou Y.
        • Li P.
        • et al.
        Diagnostic efficacy of cell block immunohistochemistry, smear cytology, and liquid-based cytology in endoscopic ultrasound-guided fine-needle aspiration of pancreatic lesions: a single-institution experience.
        PLoS One. 2014; 9e108762
        • Lee J.K.
        • Choi E.R.
        • Jang T.H.
        • et al.
        A prospective comparison of liquid-based cytology and traditional smear cytology in pancreatic endoscopic ultrasound-guided fine needle aspiration.
        Acta Cytol. 2011; 55: 401-407
        • Fontaine D.
        • Narine N.
        • Naugler C.
        Unsatisfactory rates vary between cervical cytology samples prepared using ThinPrep and SurePath platforms: a review and meta-analysis.
        BMJ Open. 2012; 2e000847
        • Yamao K.
        • Sawaki A.
        • Mizuno N.
        • et al.
        Endoscopic ultrasound-guided fine-needle aspiration biopsy (EUS-FNAB): past, present, and future.
        J Gastroenterol. 2005; 40: 1013-1023
        • Erickson R.A.
        EUS-guided FNA.
        Gastrointest Endosc. 2004; 60: 267-279
        • Yeon M.H.
        • Jeong H.S.
        • Lee H.S.
        • et al.
        Comparison of liquid-based cytology (CellPrepPlus) and conventional smears in pancreaticobiliary disease.
        Korean J Intern Med. 2018; 33: 883-892
        • Hafez N.H.
        • Shaaban H.M.
        SurePath liquid-based cytology versus conventional smears for interpretation of serous effusion fluids: a study of 104 cases.
        Egyptian J Pathol. 2018; 38: 154-161
        • Bigras G.
        • Rieder M.A.
        • Lambercy J.M.
        • et al.
        Keeping collecting device in liquid medium is mandatory to ensure optimized liquid-based cervical cytologic sampling.
        J Low Genit Tract Dis. 2003; 7: 168-174
        • Naeem R.C.
        • Goldstein D.Y.
        • Einstein M.H.
        • et al.
        SurePath specimens versus ThinPrep specimen types on the COBAS 4800 platform: high-risk HPV status and cytology correlation in an ethnically diverse Bronx population.
        Lab Med. 2017; 48: 207-213
        • Rozemeijer K.
        • Penning C.
        • Siebers A.G.
        • et al.
        Comparing SurePath, ThinPrep, and conventional cytology as primary test method: SurePath is associated with increased CIN II+ detection rates.
        Cancer Causes Control. 2016; 27: 15-25
        • Sweeney B.J.
        • Haq Z.
        • Happel J.F.
        • et al.
        Comparison of the effectiveness of two liquid-based Papanicolaou systems in the handling of adverse limiting factors, such as excessive blood.
        Cancer Cytopathol. 2006; 108: 27-31
        • Jang D.K.
        • Lee S.H.
        • Lee J.K.
        • et al.
        Comparison of cytological and histological preparations in the diagnosis of pancreatic malignancies using endoscopic ultrasound-guided fine needle aspiration.
        Hepatobiliary Pancreat Dis Int. 2017; 16: 418-423
        • Denton K.J.
        Liquid based cytology in cervical cancer screening.
        BMJ. 2007; 335: 1-2

      Linked Article

      • Optimal EUS-guided FNA cytology preparation when rapid on-site evaluation is not available
        Gastrointestinal EndoscopyVol. 91Issue 4
        • Preview
          Pancreatic adenocarcinoma is the fourth leading cause of cancer mortality in the United States, and its incidence is rising.1 EUS with FNA (EUS-FNA) has become the main technique for evaluating pancreatobiliary disorders, especially pancreatic neoplasia.2 Many patients with pancreatic cancer present at an advanced stage requiring systemic treatment. Accurate and definitive diagnosis is required before that treatment.
        • Full-Text
        • PDF