- Email: [email protected]
Does cytotechnician training influence the accuracy of EUS-guided fine-needle aspiration of pancreatic masses?
Digestive and Liver Disease 44 (2012) 311–314
Contents lists available at SciVerse ScienceDirect
Digestive and Liver Disease journal homepage: www.elsevier.com/locate/dld
Liver, Pancreas and Biliary Tract
Does cytotechnician training influence the accuracy of EUS-guided fine-needle aspiration of pancreatic masses? Maria Chiara Petrone a,∗ , Paolo Giorgio Arcidiacono a , Silvia Carrara a , Gianni Mezzi a , Claudio Doglioni b , Pier Alberto Testoni a a b
Division of Gastroenterology and Gastrointestinal Endoscopy, Vita-Salute San Raffaele University, Scientific Institute San Raffaele, Milan, Italy Pathology Unit, Vita-Salute San Raffaele University, Scientific Institute San Raffaele, Milan, Italy
a r t i c l e
i n f o
Article history: Received 7 June 2011 Accepted 1 December 2011 Available online 4 January 2012 Keywords: Endoscopic ultrasound Fine needle aspiration
a b s t r a c t Background/aim: The presence of on-site cytopathologists improves the diagnostic yield of endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) of pancreatic masses; however, on-site cytopathologists are not available to all endoscopic units. We hypothesized that experienced cytotechnicians can accurately assess whether an on-site pancreatic mass fine needle aspiration specimen is adequate. The aim of this study was to evaluate the effect of formal cytotechnician training on the diagnostic accuracy of EUS-FNA of pancreatic masses. Methods: Single-centre, prospective study. The cytotechnician made an on-site assessment of specimen adequacy with immediate evaluation of smears over a 12-month period (pre-training period) then over another 12-month period (post-training period), with a year’s intermediate training when the cytopathologist and the cytotechnician worked together in the room. The gold standard used to establish the final diagnosis was based on a non-equivocal fine needle aspiration biopsy reviewed by the same expert cytopathologist. The main outcome measurements were the cytotechnician diagnostic accuracy before and after the training period. Results: A total of 107 patients were enrolled in the pre-training period. Cytotechnician in-room adequacy was 68.2% (73/107). The diagnostic accuracy was 74.8%. The adequacy for the blind-review pathologist was 93.4% (100/107), significantly higher (p = 0.008) than the cytotechnician’s results. During the posttraining period, 95 EUS-FNA were performed and reviewed. Cytotechnician in-room adequacy was 87.4% (83/95). The diagnostic accuracy was 90.5%. The adequacy for the blinded pathologist was 95.8% (91/95), not significantly different from the cytotechnician (p = 0.23). Conclusions: An adequate training period with an expert pathologist significantly improves the cytotechnician skill in terms of judging adequacy and diagnostic accuracy. © 2012 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.
1. Introduction Endoscopic ultrasound-guided fine-needle aspiration (EUSFNA) has become a well-established procedure for tissue diagnosis in patients with suspected pancreatic masses. In tertiary centres EUS-FNA has high accuracy for diagnosing pancreatic masses, with sensitivity from 80% to 90% [1–4]. Some studies have shown that the presence of a cytopathologist in the EUS-room (on-site) further improves the diagnostic yield of EUS-FNA [5–7]. However, depending on logistics and costs, an on-site cytopathologist may not always be available and sometimes even the trained specialist performs poorly because the interpretation of pancreatic cytology specimens is challenging and there must be enough cellular
∗ Corresponding author. Tel.: +39 02 2643 2744; fax: +39 02 2643 2504. E-mail address: [email protected] (M.C. Petrone).
material to permit a definitive diagnosis. Other studies have shown that specimen adequacy and final diagnosis are directly affected more by the endosonographer’s than the cytopathologist’s experience [8]. So far no data are available on how cytopathology training influences the evaluation of specimen adequacy and diagnosis, or the cytotechnician’s yield in the diagnostic process. The aim of the present prospective study was to investigate whether specific cytotechnical training influences the accuracy of EUS-FNAs of solid pancreatic lesions. 2. Materials and methods A prospective, single-centre study was designed over a threeyear period, involving one endosonographer, one cytotechnician, and one cytopathologist. All patients referred to our tertiary university centre for EUS for the diagnosis and staging of pancreatic
1590-8658/$36.00 © 2012 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.dld.2011.12.001
312
M.C. Petrone et al. / Digestive and Liver Disease 44 (2012) 311–314
lesions were enrolled; only patients with cystic pancreatic lesions who underwent EUS-FNA for cyst fluid aspiration were excluded. The study was reviewed and approved by our institutional ethics board. All patients enrolled gave written informed consent for endoscopy and EUS-FNA, and for data management for scientific purposes. End points of the study were to compare, before and after training with an expert cytopathologist, the on-site evaluation of specimen adequacy, the diagnostic accuracy, and the number of needle passes required for the cytotechnician to be able to define the adequacy. The adequacy of the specimen was defined as the clear presence of target organ cells that allow the cytotechnician or the pathologist to obtain an accurate diagnosis.
2.1. Study design The cytotechnician made an on-site assessment of specimen adequacy with immediate evaluation of smears over a 12-month period, from March 2004 to February 2005 (pre-training period) then over another 12-month period, from March 2006 to February 2007 (post-training period), with a year’s intermediate training (March 2005–February 2006), when the cytopathologist (CD) and the cytotechnician worked together in the room and after definitive setting of the slides. The cytotechnician, at the initiation of the study, had about ten years of experience mainly in gynaecological and hepatological smears, but he had not any prior experience with an on-site interpretation of pancreatic FNA specimens. The endosonographer who performed all the FNAs (PGA) was highly experienced (with more than 500 EUS/year). During the whole three-year study, the same expert cytopathologist reviewed the samples for final adequacy and diagnosis. Patient recruitment and data analysis were limited only to the pre- and post-training periods. The cytotechnician studied was also the same.
2.2. Technique All EUS procedures were done by an experienced endosonographer (>1000 EUS-FNA at the beginning of the study) using a linear-scanning echo-endoscope (EG-3830UT, EG-3630U, FG36UX, Pentax, Hamburg, Germany). Patients were deeply sedated with propofol and the EUS was done with anaesthesiologist monitoring. Diagnostic EUS was required for evaluating and staging the primary lesion. The pancreas was scanned from both the proximal stomach and the duodenum to visualize first the body and tail, then the head and uncinate process. Lesions in the head and neck of the pancreas were sampled through the duodenal wall, and those in the body and tail through the gastric wall. The EUS-FNA technique has been described elsewhere [9]. Doppler ultrasound was used to ensure the absence of vessels. All FNAs were performed utilising a standard 25-gauge needle (Echotip, Wilson-Cook, Winstom Salem, NC). The aspirated material was totally expelled onto slides for conventional smears and prepared as previously described [9]. It was sprayed onto the slides by air using a syringe and adequacy was assessed for cell samples stained with the Diff-Quick technique. (Fig. 1) The cytotechnician reviewed the smears immediately on site to ensure the specimen was adequate. Within 24–48 h, the exact same slides were then reviewed by a single blinded cytopathologist (CD) with experience in pancreatic FNA interpretation, to judge the adequacy and give the final diagnosis. The on-site and the final cytologic diagnoses were classified as follows: positive for
Fig. 1. On-site cytological evaluation of a pancreatic sample obtained by EUS-FNA (Diff-Quick).
malignancy, suspicious for malignancy, atypical cells, benign or reactive process and inadequate. 2.3. Statistical analyses Descriptive statistics are given as frequencies and percentages for categorical data and means with standard deviation (sd) for continuous data. Student’s t-test was used for continuous variables, and the two-sided chi-squared Pearson’s test. In all analyses, a p value less than 0.05 was considered significant. For this study, the agreement between the cytotechnician and final pathologist interpretation was assessed by the “kappa” statistic. k statistics are widely used and accepted coefficients that provide a measure of observer agreement accounting for agreement other than that which occurs by chance alone. A kappa value of ≤0.20 was considered as poor agreement, 0.21–0.40 as fair, 0.41–0.60 as moderate, 0.61–0.80 as good, and ≥0.81 as an excellent agreement. 3. Results A total of 107 patients were enrolled in the pre-training period. Cytotechnician in-room adequacy was 68.2% (73/107), with sensitivity and specificity of 71.4% and 93.7%, respectively. The positive predictive value (PPV) was 98.4% and negative predictive value (NPV) 36.5%. The diagnostic accuracy was 74.8%. The adequacy for the blind-review pathologist was 93.4% (100/107), significantly higher (p = 0.008) than the cytotechnician’s results. Overall, in the training period 280 pancreatic mass EUS-FNAs were done. During the post-training period 95 pancreatic EUS-FNAs were done and reviewed. Cytotechnician in-room adequacy was 87.4% (83/95). Sensitivity and specificity were 89.2% and 100%, respectively. PPV and NPV were 100% and 57.1%, respectively. The diagnostic accuracy was 90.5%. The adequacy for the blinded pathologist was 95.8% (91/95), not significantly different from the cytotechnician (p = 0.23). In the post-training period, all the 63 lesions initially reported as malignant by the cytotechnician were confirmed on final review by the cytopathologist. Of these, all 20 patients who were subsequently resected, the cytological diagnosis was confirmed at the histology examination. None of the lesions initially recognized as suspicious for malignancy was downgraded to benign on final interpretation. Of the ten lesions initially reported as atypical by the cytotechnician, nine remained atypical (90%) and one was
M.C. Petrone et al. / Digestive and Liver Disease 44 (2012) 311–314
313
Table 1 Pre-training period: on-site cytotechnician’s and pathologist’s final diagnoses of 107 pancreatic masses. EUS-FNA final cytologic diagnosis
EUS-FNA on-site cytology Malignant Suspicious Atypical Benign Inadequate Total
Malignant
Suspicious
Atypical
Benign
Inadequate
Total
61 4 – 2 6 73
– – 1 – – 1
– – – – – –
3 1 1 11 17 33
– – – – – –
64 5 2 13 23 107
EUS-FNA, endoscopic ultrasound-fine needle aspiration.
Table 2 Post-training period: on-site cytotechnician’s and pathologist’s final diagnoses of 95 solid pancreatic masses. EUS-FNA final cytologic diagnosis
EUS-FNA on-site cytology Malignant Suspicious Atypical Benign Inadequate Total
Malignant
Suspicious
Atypical
Benign
Inadequate
Total
63 1 – 3 1 68
– 2 – – – 2
– – 9 1 – 10
– – – 10 – 10
– – 1 1 3 5
63 3 10 15 4 95
EUS-FNA, endoscopic ultrasound-fine needle aspiration.
Table 3 Characteristics of pancreatic solid lesions.
Lesion localization Head Body Tail Mean size (mean ± sd)
Pre-training period (n = 107)
Post-training period (n = 95)
p value
77 23 7 29.7 ± 12.6 mm
71 20 4 30.7 ± 13.5 mm
ns ns ns ns
considered inadequate. Out of the 15 benign lesions interpreted by on-site evaluation, ten (66.6%) were reported as benign in the final cytology report; three were upgraded to malignant, one atypical and one inadequate. Of the four lesions considered inadequate at on-site evaluation, one was interpreted as malignant by final cytology; the other three remained inadequate. Only 8.4% (8/95) of the lesions had a different final cytology interpretation from the one given on-site. Tables 1 and 2 compare the initial cytotechnician’s and final cytopathologist’s diagnosis in the two periods. Differences in tumour size and location were not significant (Table 3). Overall, the in-room diagnostic accuracy in the post-training period was significantly higher than during pre-training (91.6% vs. 74.8%; p = 0.003). The adequacy of the blinded pathologist reviewer was significantly higher (p = 0.008) than the cytotechnician’s judgments in the pre-training period, but not after training (p = 0.23).
Table 4 Diagnostic value of endoscopic ultrasound-fine needle aspiration in pre- and posttraining periods.
Sensitivity Specificity PPV NPV Accuracy
Pre-training period n = 107 Mean (sd)
Post-training period n = 95 Mean (sd)
p value
71.4% (65/91) 93.7% (15/16) 98.4% (65/66) 36.5% (15/41) 74.8% (80/107)
89.2% (75/84) 100% (11/11) 100% (75/75) 57.1% (11/20) 90.5% (86/95)
0.003* 0.39 0.28 0.17 0.003*
PPV, positive predictive value; NPV, negative predictive value. * Statistically significant.
Table 4 summarizes the diagnostic accuracy for benign or malignant masses. The numbers of passes needed to obtain adequate samples and to diagnose malignancy were significantly lower (p = 0.01) after training than before; the mean needle passes per procedure were, respectively, 3.7 ± 1.1 and 2.5 ± 0.7. The agreement between the cytotechnician and final pathologist interpretation was moderate before the training period (k = 0.44; 95% CI, 0.35–0.53), whilst we observed that there was a good agreement between the on-site and final cytology interpretation results after the training period (k = 0.79; 95% CI, 0.68–0.94). 4. Discussion EUS-FNA has become a powerful tool in the diagnosis and the staging of pancreatic masses. Rapid on-site evaluation of cytologic specimens for adequacy improves the accuracy and reduces the need for repeated examination by ensuring sufficient material before terminating the procedure. Chang et al. [5], in a multicentre study with EUS-FNA of the pancreas, found that the diagnostic yield of the procedure increased in terms of greater sensitivity, accuracy, and a higher percentage of adequate specimens at two centres where on-site cytopathology interpretation was available. Klapman et al. [6], in a prospective series, analysed EUSguided FNA results from two different hospitals, with and without attending cytopathologist. On-site cytopathology interpretation during EUS-guided FNA had a significant clinical impact, increasing the diagnostic yield of the FNA. Eloubeidi et al. [7] prospectively evaluated consecutive EUS-FNA specimens obtained by a
314
M.C. Petrone et al. / Digestive and Liver Disease 44 (2012) 311–314
single endosonographer with a cytopathologist present. They found excellent agreement between rapid on-site evaluation and final cytology interpretation, compared using the “kappa” statistic. In a study of 109 EUS-FNA of the pancreas the absence of a cytopathologist in the endoscopy room increased the number of passes, with a 10–15% reduction in the rate of definitive cytologic diagnoses [10]. Iglesias-Garcia et al. evaluated the influence of on-site cytopathological interpretation on the diagnostic yield of EUSguided FNA for the diagnosis of pancreatic masses comparing in an unselected series of consecutive patients. The presence of on-site cytopathologist was associated with a significantly lower number of inadequate sample and a significantly higher diagnostic sensitivity when compared with the samples obtained without an on-site cytopathologist [11]. Although many experts advocate immediate on-site tissue sample evaluation by a cytopathologist to optimize the diagnostic yield, the practice of on-site cytology interpretation varies in different pathology units. The cytopathologist is not always available and frequently, in an attempt to reduce costs related to the on-site procedure, a cytotechnician ensures specimen adequacy. In addition, studies specifically designed with the aim of evaluating the influence of a specific cytotechnician training on diagnostic accuracy are scarse. No data are available on the possibility of calling in a cytotechnician in place of a cytopathologist not only for quantitative determination of specimen adequacy but also to identify the type of cells and, if possible, give a preliminary diagnosis as well. In our study we found that a cytotechnician could reach a high level of skill in terms of judging adequacy and diagnostic accuracy, after an adequate training period alongside an expert cytopathologist. The cytotechnician’s adequacy in the pre-training period was 68.2%, and rose to 91.6% after the training period – significantly higher. Moreover, after the training, the difference in adequacy judgments between the cytotechnician and the expert cytopathologist was not statistically significant. The cytotechnician’s diagnostic accuracy in the post-training period was 90.5%. This suggests that an adequate training period with an expert cytopathologist can significantly improve a cytotechnician’s accuracy in on-site EUS-FNA procedures and is associated with a significant lower number of needle passes. In this study an adequate sample was obtained with just one or two passes in the majority of patients when samples were evaluated on-site by the experienced cytotechnician, compared with three to four passes usually performed when the cytotechnician was not trained. A major strength of the study was that the same experienced endoscopist performed all procedures in both periods, allowing us to exclude the endosonographer’s ability and experience as a reason for the difference in procedure results. We also found that when the EUS-FNA was done by an expert endosonographer the differences in adequacy and diagnostic rates are completely dependent on the cytotechnician’s skill. In our daily practice, cytotechnicians regularly screen samples and confirm their adequacy on site,
formulating a preliminary diagnosis before the procedure is terminated. Weaknesses of this study include the lack of histology as “gold standard”, since the majority of patients had non-resectable disease. In all 20 patients (21%) who were subsequently resected, the cytological diagnosis was confirmed at the histology examination. Moreover, a major limitation of the manuscript is that although we can assume that the reason of the cytotechnician’s improved rate of adequacy and diagnostic determination was do to training, it would have been better to have another cytotechnician who started at the same time, and compare the first and third year results of the untrained cytotechnician to the trained cytotechnician. In conclusion, we believe that formally trained cytotechnicians can make an on-site interpretation of cytologic adequacy; they are first-line interpreters and their level of performance and contribution have a direct impact on the efficiency of the procedure. An adequate training period with an expert cytopathologist boosts the cytotechnician’s accuracy in EUS-FNA procedures. This influences the number of needle passes, and the diagnostic yield of the procedures. Conflict of interest statement No conflict of interests exist. References [1] Bhutani MS, Hawes RH, Baron PL, et al. Endoscopic ultrasound-guided fine needle aspiration of malignant pancreatic lesions. Endoscopy 1997;29:854–8. [2] Faigel DO, Ginsberg GG, Bentz JS, et al. Endoscopic ultrasound-guided fine needle aspiration biopsy of the pancreas in cancer patients with pancreatic lesions. J Clin Oncol 1997;15:1439–43. [3] Chang KJ, Nguyen P, Erickson RA, et al. The clinical utility of endoscopic ultrasound-guided fine-needle aspiration in the diagnosis and staging of pancreatic carcinoma. Gastrointest Endosc 1997;45:387–93. [4] Harewood GC, Wiersema MJ. Endosonography-guided fine needle aspiration biopsy in the diagnosis of solid pancreatic masses. Am J Gastroenterol 2002;97:1386–91. [5] Chang KJ, Wiersema M, Giovannini M, et al. Multi-center experience with endoscopic ultrasound (EUS) guided fine-needle aspiration (FNA) of the pancreas. Gastrointest Endosc 1996;43:S49. [6] Klapman JB, Logrono R, Dye CE, et al. Clinical impact of on-site cytopathology interpretation on endoscopic ultrasound-guided fine-needle aspiration. Am J Gastroenterol 2003;98:1289–94. [7] Eloubeidi MA, Tamhane A, Jhala N, et al. Agreement between rapid in-site and final cytologic interpretations of EUS-guided FNA specimens: implications for the endosonographer and patient management. Am J Gastroenterol 2006;101:2841–7. [8] Harewood GC, Wiersema LM, Halling AC, et al. Influence of EUS training and pathology interpretation on accuracy of EUS-guided fine needle aspiration of pancreatic masses. Gastrointest Endosc 2002;55:669–73. [9] Jhala NC, Jhala DN, Chhieng DC, et al. Endoscopic ultrasound-guided fineneedle aspiration. A cytopathologist’s perspective. Am J Clin Pathol 2003;120: 351–67. [10] Erickson RA, Sayage-Rabie L, Beissner RS. Factors predicting the number of EUS-guided fine-needle passes for diagnosis of pancreatic malignancies. Gastrointest Endosc 2000;51:184–90. [11] Iglesias-Garcia J, Dominguez-Munoz JE, Abdulkader I, et al. Influence of on-site cytopathology evaluation on the diagnostic accuracy of endoscopic ultrasoundguided fine needle aspiration (EUS-FNA) of solid pancreatic masses. Am J Gastroenterol 2010;106(9):1705–10.
Contents lists available at SciVerse ScienceDirect
Digestive and Liver Disease journal homepage: www.elsevier.com/locate/dld
Liver, Pancreas and Biliary Tract
Does cytotechnician training influence the accuracy of EUS-guided fine-needle aspiration of pancreatic masses? Maria Chiara Petrone a,∗ , Paolo Giorgio Arcidiacono a , Silvia Carrara a , Gianni Mezzi a , Claudio Doglioni b , Pier Alberto Testoni a a b
Division of Gastroenterology and Gastrointestinal Endoscopy, Vita-Salute San Raffaele University, Scientific Institute San Raffaele, Milan, Italy Pathology Unit, Vita-Salute San Raffaele University, Scientific Institute San Raffaele, Milan, Italy
a r t i c l e
i n f o
Article history: Received 7 June 2011 Accepted 1 December 2011 Available online 4 January 2012 Keywords: Endoscopic ultrasound Fine needle aspiration
a b s t r a c t Background/aim: The presence of on-site cytopathologists improves the diagnostic yield of endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) of pancreatic masses; however, on-site cytopathologists are not available to all endoscopic units. We hypothesized that experienced cytotechnicians can accurately assess whether an on-site pancreatic mass fine needle aspiration specimen is adequate. The aim of this study was to evaluate the effect of formal cytotechnician training on the diagnostic accuracy of EUS-FNA of pancreatic masses. Methods: Single-centre, prospective study. The cytotechnician made an on-site assessment of specimen adequacy with immediate evaluation of smears over a 12-month period (pre-training period) then over another 12-month period (post-training period), with a year’s intermediate training when the cytopathologist and the cytotechnician worked together in the room. The gold standard used to establish the final diagnosis was based on a non-equivocal fine needle aspiration biopsy reviewed by the same expert cytopathologist. The main outcome measurements were the cytotechnician diagnostic accuracy before and after the training period. Results: A total of 107 patients were enrolled in the pre-training period. Cytotechnician in-room adequacy was 68.2% (73/107). The diagnostic accuracy was 74.8%. The adequacy for the blind-review pathologist was 93.4% (100/107), significantly higher (p = 0.008) than the cytotechnician’s results. During the posttraining period, 95 EUS-FNA were performed and reviewed. Cytotechnician in-room adequacy was 87.4% (83/95). The diagnostic accuracy was 90.5%. The adequacy for the blinded pathologist was 95.8% (91/95), not significantly different from the cytotechnician (p = 0.23). Conclusions: An adequate training period with an expert pathologist significantly improves the cytotechnician skill in terms of judging adequacy and diagnostic accuracy. © 2012 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.
1. Introduction Endoscopic ultrasound-guided fine-needle aspiration (EUSFNA) has become a well-established procedure for tissue diagnosis in patients with suspected pancreatic masses. In tertiary centres EUS-FNA has high accuracy for diagnosing pancreatic masses, with sensitivity from 80% to 90% [1–4]. Some studies have shown that the presence of a cytopathologist in the EUS-room (on-site) further improves the diagnostic yield of EUS-FNA [5–7]. However, depending on logistics and costs, an on-site cytopathologist may not always be available and sometimes even the trained specialist performs poorly because the interpretation of pancreatic cytology specimens is challenging and there must be enough cellular
∗ Corresponding author. Tel.: +39 02 2643 2744; fax: +39 02 2643 2504. E-mail address: [email protected] (M.C. Petrone).
material to permit a definitive diagnosis. Other studies have shown that specimen adequacy and final diagnosis are directly affected more by the endosonographer’s than the cytopathologist’s experience [8]. So far no data are available on how cytopathology training influences the evaluation of specimen adequacy and diagnosis, or the cytotechnician’s yield in the diagnostic process. The aim of the present prospective study was to investigate whether specific cytotechnical training influences the accuracy of EUS-FNAs of solid pancreatic lesions. 2. Materials and methods A prospective, single-centre study was designed over a threeyear period, involving one endosonographer, one cytotechnician, and one cytopathologist. All patients referred to our tertiary university centre for EUS for the diagnosis and staging of pancreatic
1590-8658/$36.00 © 2012 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.dld.2011.12.001
312
M.C. Petrone et al. / Digestive and Liver Disease 44 (2012) 311–314
lesions were enrolled; only patients with cystic pancreatic lesions who underwent EUS-FNA for cyst fluid aspiration were excluded. The study was reviewed and approved by our institutional ethics board. All patients enrolled gave written informed consent for endoscopy and EUS-FNA, and for data management for scientific purposes. End points of the study were to compare, before and after training with an expert cytopathologist, the on-site evaluation of specimen adequacy, the diagnostic accuracy, and the number of needle passes required for the cytotechnician to be able to define the adequacy. The adequacy of the specimen was defined as the clear presence of target organ cells that allow the cytotechnician or the pathologist to obtain an accurate diagnosis.
2.1. Study design The cytotechnician made an on-site assessment of specimen adequacy with immediate evaluation of smears over a 12-month period, from March 2004 to February 2005 (pre-training period) then over another 12-month period, from March 2006 to February 2007 (post-training period), with a year’s intermediate training (March 2005–February 2006), when the cytopathologist (CD) and the cytotechnician worked together in the room and after definitive setting of the slides. The cytotechnician, at the initiation of the study, had about ten years of experience mainly in gynaecological and hepatological smears, but he had not any prior experience with an on-site interpretation of pancreatic FNA specimens. The endosonographer who performed all the FNAs (PGA) was highly experienced (with more than 500 EUS/year). During the whole three-year study, the same expert cytopathologist reviewed the samples for final adequacy and diagnosis. Patient recruitment and data analysis were limited only to the pre- and post-training periods. The cytotechnician studied was also the same.
2.2. Technique All EUS procedures were done by an experienced endosonographer (>1000 EUS-FNA at the beginning of the study) using a linear-scanning echo-endoscope (EG-3830UT, EG-3630U, FG36UX, Pentax, Hamburg, Germany). Patients were deeply sedated with propofol and the EUS was done with anaesthesiologist monitoring. Diagnostic EUS was required for evaluating and staging the primary lesion. The pancreas was scanned from both the proximal stomach and the duodenum to visualize first the body and tail, then the head and uncinate process. Lesions in the head and neck of the pancreas were sampled through the duodenal wall, and those in the body and tail through the gastric wall. The EUS-FNA technique has been described elsewhere [9]. Doppler ultrasound was used to ensure the absence of vessels. All FNAs were performed utilising a standard 25-gauge needle (Echotip, Wilson-Cook, Winstom Salem, NC). The aspirated material was totally expelled onto slides for conventional smears and prepared as previously described [9]. It was sprayed onto the slides by air using a syringe and adequacy was assessed for cell samples stained with the Diff-Quick technique. (Fig. 1) The cytotechnician reviewed the smears immediately on site to ensure the specimen was adequate. Within 24–48 h, the exact same slides were then reviewed by a single blinded cytopathologist (CD) with experience in pancreatic FNA interpretation, to judge the adequacy and give the final diagnosis. The on-site and the final cytologic diagnoses were classified as follows: positive for
Fig. 1. On-site cytological evaluation of a pancreatic sample obtained by EUS-FNA (Diff-Quick).
malignancy, suspicious for malignancy, atypical cells, benign or reactive process and inadequate. 2.3. Statistical analyses Descriptive statistics are given as frequencies and percentages for categorical data and means with standard deviation (sd) for continuous data. Student’s t-test was used for continuous variables, and the two-sided chi-squared Pearson’s test. In all analyses, a p value less than 0.05 was considered significant. For this study, the agreement between the cytotechnician and final pathologist interpretation was assessed by the “kappa” statistic. k statistics are widely used and accepted coefficients that provide a measure of observer agreement accounting for agreement other than that which occurs by chance alone. A kappa value of ≤0.20 was considered as poor agreement, 0.21–0.40 as fair, 0.41–0.60 as moderate, 0.61–0.80 as good, and ≥0.81 as an excellent agreement. 3. Results A total of 107 patients were enrolled in the pre-training period. Cytotechnician in-room adequacy was 68.2% (73/107), with sensitivity and specificity of 71.4% and 93.7%, respectively. The positive predictive value (PPV) was 98.4% and negative predictive value (NPV) 36.5%. The diagnostic accuracy was 74.8%. The adequacy for the blind-review pathologist was 93.4% (100/107), significantly higher (p = 0.008) than the cytotechnician’s results. Overall, in the training period 280 pancreatic mass EUS-FNAs were done. During the post-training period 95 pancreatic EUS-FNAs were done and reviewed. Cytotechnician in-room adequacy was 87.4% (83/95). Sensitivity and specificity were 89.2% and 100%, respectively. PPV and NPV were 100% and 57.1%, respectively. The diagnostic accuracy was 90.5%. The adequacy for the blinded pathologist was 95.8% (91/95), not significantly different from the cytotechnician (p = 0.23). In the post-training period, all the 63 lesions initially reported as malignant by the cytotechnician were confirmed on final review by the cytopathologist. Of these, all 20 patients who were subsequently resected, the cytological diagnosis was confirmed at the histology examination. None of the lesions initially recognized as suspicious for malignancy was downgraded to benign on final interpretation. Of the ten lesions initially reported as atypical by the cytotechnician, nine remained atypical (90%) and one was
M.C. Petrone et al. / Digestive and Liver Disease 44 (2012) 311–314
313
Table 1 Pre-training period: on-site cytotechnician’s and pathologist’s final diagnoses of 107 pancreatic masses. EUS-FNA final cytologic diagnosis
EUS-FNA on-site cytology Malignant Suspicious Atypical Benign Inadequate Total
Malignant
Suspicious
Atypical
Benign
Inadequate
Total
61 4 – 2 6 73
– – 1 – – 1
– – – – – –
3 1 1 11 17 33
– – – – – –
64 5 2 13 23 107
EUS-FNA, endoscopic ultrasound-fine needle aspiration.
Table 2 Post-training period: on-site cytotechnician’s and pathologist’s final diagnoses of 95 solid pancreatic masses. EUS-FNA final cytologic diagnosis
EUS-FNA on-site cytology Malignant Suspicious Atypical Benign Inadequate Total
Malignant
Suspicious
Atypical
Benign
Inadequate
Total
63 1 – 3 1 68
– 2 – – – 2
– – 9 1 – 10
– – – 10 – 10
– – 1 1 3 5
63 3 10 15 4 95
EUS-FNA, endoscopic ultrasound-fine needle aspiration.
Table 3 Characteristics of pancreatic solid lesions.
Lesion localization Head Body Tail Mean size (mean ± sd)
Pre-training period (n = 107)
Post-training period (n = 95)
p value
77 23 7 29.7 ± 12.6 mm
71 20 4 30.7 ± 13.5 mm
ns ns ns ns
considered inadequate. Out of the 15 benign lesions interpreted by on-site evaluation, ten (66.6%) were reported as benign in the final cytology report; three were upgraded to malignant, one atypical and one inadequate. Of the four lesions considered inadequate at on-site evaluation, one was interpreted as malignant by final cytology; the other three remained inadequate. Only 8.4% (8/95) of the lesions had a different final cytology interpretation from the one given on-site. Tables 1 and 2 compare the initial cytotechnician’s and final cytopathologist’s diagnosis in the two periods. Differences in tumour size and location were not significant (Table 3). Overall, the in-room diagnostic accuracy in the post-training period was significantly higher than during pre-training (91.6% vs. 74.8%; p = 0.003). The adequacy of the blinded pathologist reviewer was significantly higher (p = 0.008) than the cytotechnician’s judgments in the pre-training period, but not after training (p = 0.23).
Table 4 Diagnostic value of endoscopic ultrasound-fine needle aspiration in pre- and posttraining periods.
Sensitivity Specificity PPV NPV Accuracy
Pre-training period n = 107 Mean (sd)
Post-training period n = 95 Mean (sd)
p value
71.4% (65/91) 93.7% (15/16) 98.4% (65/66) 36.5% (15/41) 74.8% (80/107)
89.2% (75/84) 100% (11/11) 100% (75/75) 57.1% (11/20) 90.5% (86/95)
0.003* 0.39 0.28 0.17 0.003*
PPV, positive predictive value; NPV, negative predictive value. * Statistically significant.
Table 4 summarizes the diagnostic accuracy for benign or malignant masses. The numbers of passes needed to obtain adequate samples and to diagnose malignancy were significantly lower (p = 0.01) after training than before; the mean needle passes per procedure were, respectively, 3.7 ± 1.1 and 2.5 ± 0.7. The agreement between the cytotechnician and final pathologist interpretation was moderate before the training period (k = 0.44; 95% CI, 0.35–0.53), whilst we observed that there was a good agreement between the on-site and final cytology interpretation results after the training period (k = 0.79; 95% CI, 0.68–0.94). 4. Discussion EUS-FNA has become a powerful tool in the diagnosis and the staging of pancreatic masses. Rapid on-site evaluation of cytologic specimens for adequacy improves the accuracy and reduces the need for repeated examination by ensuring sufficient material before terminating the procedure. Chang et al. [5], in a multicentre study with EUS-FNA of the pancreas, found that the diagnostic yield of the procedure increased in terms of greater sensitivity, accuracy, and a higher percentage of adequate specimens at two centres where on-site cytopathology interpretation was available. Klapman et al. [6], in a prospective series, analysed EUSguided FNA results from two different hospitals, with and without attending cytopathologist. On-site cytopathology interpretation during EUS-guided FNA had a significant clinical impact, increasing the diagnostic yield of the FNA. Eloubeidi et al. [7] prospectively evaluated consecutive EUS-FNA specimens obtained by a
314
M.C. Petrone et al. / Digestive and Liver Disease 44 (2012) 311–314
single endosonographer with a cytopathologist present. They found excellent agreement between rapid on-site evaluation and final cytology interpretation, compared using the “kappa” statistic. In a study of 109 EUS-FNA of the pancreas the absence of a cytopathologist in the endoscopy room increased the number of passes, with a 10–15% reduction in the rate of definitive cytologic diagnoses [10]. Iglesias-Garcia et al. evaluated the influence of on-site cytopathological interpretation on the diagnostic yield of EUSguided FNA for the diagnosis of pancreatic masses comparing in an unselected series of consecutive patients. The presence of on-site cytopathologist was associated with a significantly lower number of inadequate sample and a significantly higher diagnostic sensitivity when compared with the samples obtained without an on-site cytopathologist [11]. Although many experts advocate immediate on-site tissue sample evaluation by a cytopathologist to optimize the diagnostic yield, the practice of on-site cytology interpretation varies in different pathology units. The cytopathologist is not always available and frequently, in an attempt to reduce costs related to the on-site procedure, a cytotechnician ensures specimen adequacy. In addition, studies specifically designed with the aim of evaluating the influence of a specific cytotechnician training on diagnostic accuracy are scarse. No data are available on the possibility of calling in a cytotechnician in place of a cytopathologist not only for quantitative determination of specimen adequacy but also to identify the type of cells and, if possible, give a preliminary diagnosis as well. In our study we found that a cytotechnician could reach a high level of skill in terms of judging adequacy and diagnostic accuracy, after an adequate training period alongside an expert cytopathologist. The cytotechnician’s adequacy in the pre-training period was 68.2%, and rose to 91.6% after the training period – significantly higher. Moreover, after the training, the difference in adequacy judgments between the cytotechnician and the expert cytopathologist was not statistically significant. The cytotechnician’s diagnostic accuracy in the post-training period was 90.5%. This suggests that an adequate training period with an expert cytopathologist can significantly improve a cytotechnician’s accuracy in on-site EUS-FNA procedures and is associated with a significant lower number of needle passes. In this study an adequate sample was obtained with just one or two passes in the majority of patients when samples were evaluated on-site by the experienced cytotechnician, compared with three to four passes usually performed when the cytotechnician was not trained. A major strength of the study was that the same experienced endoscopist performed all procedures in both periods, allowing us to exclude the endosonographer’s ability and experience as a reason for the difference in procedure results. We also found that when the EUS-FNA was done by an expert endosonographer the differences in adequacy and diagnostic rates are completely dependent on the cytotechnician’s skill. In our daily practice, cytotechnicians regularly screen samples and confirm their adequacy on site,
formulating a preliminary diagnosis before the procedure is terminated. Weaknesses of this study include the lack of histology as “gold standard”, since the majority of patients had non-resectable disease. In all 20 patients (21%) who were subsequently resected, the cytological diagnosis was confirmed at the histology examination. Moreover, a major limitation of the manuscript is that although we can assume that the reason of the cytotechnician’s improved rate of adequacy and diagnostic determination was do to training, it would have been better to have another cytotechnician who started at the same time, and compare the first and third year results of the untrained cytotechnician to the trained cytotechnician. In conclusion, we believe that formally trained cytotechnicians can make an on-site interpretation of cytologic adequacy; they are first-line interpreters and their level of performance and contribution have a direct impact on the efficiency of the procedure. An adequate training period with an expert cytopathologist boosts the cytotechnician’s accuracy in EUS-FNA procedures. This influences the number of needle passes, and the diagnostic yield of the procedures. Conflict of interest statement No conflict of interests exist. References [1] Bhutani MS, Hawes RH, Baron PL, et al. Endoscopic ultrasound-guided fine needle aspiration of malignant pancreatic lesions. Endoscopy 1997;29:854–8. [2] Faigel DO, Ginsberg GG, Bentz JS, et al. Endoscopic ultrasound-guided fine needle aspiration biopsy of the pancreas in cancer patients with pancreatic lesions. J Clin Oncol 1997;15:1439–43. [3] Chang KJ, Nguyen P, Erickson RA, et al. The clinical utility of endoscopic ultrasound-guided fine-needle aspiration in the diagnosis and staging of pancreatic carcinoma. Gastrointest Endosc 1997;45:387–93. [4] Harewood GC, Wiersema MJ. Endosonography-guided fine needle aspiration biopsy in the diagnosis of solid pancreatic masses. Am J Gastroenterol 2002;97:1386–91. [5] Chang KJ, Wiersema M, Giovannini M, et al. Multi-center experience with endoscopic ultrasound (EUS) guided fine-needle aspiration (FNA) of the pancreas. Gastrointest Endosc 1996;43:S49. [6] Klapman JB, Logrono R, Dye CE, et al. Clinical impact of on-site cytopathology interpretation on endoscopic ultrasound-guided fine-needle aspiration. Am J Gastroenterol 2003;98:1289–94. [7] Eloubeidi MA, Tamhane A, Jhala N, et al. Agreement between rapid in-site and final cytologic interpretations of EUS-guided FNA specimens: implications for the endosonographer and patient management. Am J Gastroenterol 2006;101:2841–7. [8] Harewood GC, Wiersema LM, Halling AC, et al. Influence of EUS training and pathology interpretation on accuracy of EUS-guided fine needle aspiration of pancreatic masses. Gastrointest Endosc 2002;55:669–73. [9] Jhala NC, Jhala DN, Chhieng DC, et al. Endoscopic ultrasound-guided fineneedle aspiration. A cytopathologist’s perspective. Am J Clin Pathol 2003;120: 351–67. [10] Erickson RA, Sayage-Rabie L, Beissner RS. Factors predicting the number of EUS-guided fine-needle passes for diagnosis of pancreatic malignancies. Gastrointest Endosc 2000;51:184–90. [11] Iglesias-Garcia J, Dominguez-Munoz JE, Abdulkader I, et al. Influence of on-site cytopathology evaluation on the diagnostic accuracy of endoscopic ultrasoundguided fine needle aspiration (EUS-FNA) of solid pancreatic masses. Am J Gastroenterol 2010;106(9):1705–10.