Introduction
Endoscopic ultrasound (EUS)-guided fine needle aspiration (FNA) is a widely available,
commonly used, and effective modality for the evaluation of various gastrointestinal
and peri-intestinal masses including pancreatic, submucosal, and lymphatic lesions
[1]
[2]
[3]. Despite the diagnostic utility of EUS-guided FNA in the evaluation of mass lesions,
there are several limitations to the procedure. The diagnostic yield of EUS-FNA is
highly variable and influenced by the presence of an on-site cytopathologist. For
example, EUS-FNA of pancreatic lesions has a diagnostic accuracy ranging from 78 %
to 95 % [4], but these rates have been reported to be even lower for other targets [5]
[6]. Inadequate specimens are obtained in as many as 29 % of patients who undergo EUS-FNA
without immediate review by a bedside cytopathologist [7]. On-site cytopathological evaluation of FNA samples significantly decreases the
number of inadequate samples as well as the number of needle passes needed [8]
[9]. Despite these advantages, many institutions lack immediate on-site interpretation
by a cytopathologist during EUS-FNA.
Another limitation of EUS-FNA is the inherent inadequacies of cytology itself. The
absence of tissue architecture with FNA makes it difficult to diagnose stromal tumors
and lymphomas [10]
[11]. More recently, studies have examined the utility of performing EUS-guided core
biopsies as a way to overcome such limitations [12]. In the last decade, many studies have been published comparing the diagnostic yield
of EUS-FNA (without on-site cytopathology) and EUS core needle biopsy for various
gastrointestinal lesions. To our knowledge, there has not been a study comparing EUS-guided
core biopsy and EUS-FNA utilizing immediate bedside cytopathologist review. We hypothesize
that the diagnostic yield of EUS core biopsy is comparable to the diagnostic yield
of EUS-FNA combined with on-site cytopathology.
Methods
Study design
This study was conducted at a tertiary referral medical center. With the aim of increasing
diagnostic yield, all patients who were referred for EUS-guided biopsy at the Michael
E. DeBakey VA Medical Center (MEDVAMC) between October 2011 and January 2013 were
considered for both FNA and core biopsy during the same endoscopic session. Core biopsy
was not performed if patients had cystic lesions, small lesions (defined as < 1 cm)
legions with overlying vascular structures that precluded safe intervention, or if
the advanced endoscopist performing the procedure felt the risks of multiple biopsies
outweighed the benefits. Data for this study was prospectively collected as part of
a database that includes all EUS procedures performed in our endoscopy unit. Informed
consent was obtained from each patient. The protocol was approved by the Baylor College
of Medicine IRB and the MEDVAMC Research and Development Committee.
EUS-FNA and core biopsy
In this study, back-to-back EUS-FNA and core biopsy were performed on all patients
during the same endoscopic session. All procedures were performed by a single, experienced
advanced endoscopist (YS). YS has performed over 1000 cases of EUS and EUS-FNA. A
gastroenterology fellow partially assisted in some of these cases. The mass lesions
were identified using endoscopic ultrasound and sampled via two core biopsies (Cook
22-gauge Pro-Core core biopsy needle, Cook Endoscopy Inc; Limerick, Ireland) followed
by two FNA passes (Cook 22-gauge FNA needle, Cook Endoscopy Inc; Limerick, Ireland).
Suction was used to enhance cell capture. A bedside attending cytopathologist was
present during the procedure and evaluated all FNA samples. A minimum of two FNA passes
were made and more passes were obtained as needed based on the bedside cytopathologist’s
assessment of the initial FNA samples. Core biopsies were processed and evaluated
by a pathologist (LG) blinded to the FNA results. All FNA and core biopsy specimens
felt to be malignant were reviewed and confirmed by a second pathologist. LG has over
25 years of experience as a cytopathologist. She is a consultant cytopathologist for
the entire VA system. At the completion of the procedure, patients were monitored
in the post-anesthesia care unit for adverse events. Adverse events occurring after
the immediate post-procedure period were determined based on chart review.
Outcomes
The primary objective of this study was to determine the diagnostic accuracy of EUS-guided
core biopsy when compared with the accuracy of the gold standard, EUS-FNA with a bedside
cytopathologist. Accuracy was defined as the percentage of specimens in which the
biopsy diagnosis was consistent with the final diagnosis (i. e. sum of true positives
and true negatives). The final diagnosis was determined based on a combination of
surgical pathology; biopsy of primary tumor or metastatic lesions; serial imaging
including computed tomography (CT), magnetic resonance imaging (MRI), X-ray, ultrasound,
and positron emission tomography (PET) imaging; labs such as tumor markers; and clinical
course. For example, if pathological diagnosis could not be obtained after multiple
attempts with various modalities, metastatic or rapidly growing lesions were considered
to be malignant. Patients were monitored for at least 1 year following their EUS procedure.
Secondary outcomes examined in this study included procedure time and adverse events.
Procedure time was defined as the time from insertion of the endoscope to removal.
Data for secondary outcomes were collected based on a review of anesthesia, post-anesthesia
care unit, and electronic medical records.
Statistical analysis
The diagnostic accuracy, sensitivity, and specificity were calculated for both techniques.
Diagnostic accuracy (defined as true positives and true negatives) was compared using
the Chi-squared test. Agreement in diagnostic yield of EUS-guided core biopsy and
EUS-FNA with bedside cytopathologist was assessed using the kappa statistic. A multivariate
analysis was conducted to examine potential predictors of an accurate diagnosis. Data
analysis was performed using JMP 7 software (SAS Institute, Cary, NC).
Results
During the period between October 2011 and January 2013, 45 patients were referred
for EUS-guided biopsy at the MEDVAMC. Both interventions could not be performed on
19 of the 45 patients for the following reasons: seven cystic lesions, five lesions
smaller than 1 cm, four masses not seen during EUS, and three which were technically
difficult to sample. Twenty-six patients were included in the final analysis. Nineteen
(73.1 %) of the sampled lesions were pancreatic masses. Extrapancreatic lesions included
peripancreatic lymph nodes, gastric lesions, para-aortic lymph nodes, mediastinal
masses, and liver lesions. Patient demographics and other relevant clinical characteristics
are provided in [Table 1].
Table 1
Patient characteristics.
|
Characteristics
|
|
|
Age, mean ± SD, years
|
66.8 ± 8.9
|
|
Sex, male, n (%)
|
25 (96.2)
|
|
Race, Non-Hispanic White, n (%)
|
16 (61.5)
|
|
Site of lesion
|
|
|
Pancreas, n (%)
|
19 (73.1)
|
|
Other, n (%)
|
7 (26.9)
|
Results from the procedure are displayed in [Table 2]. The mean number of passes was 3.2 for FNA. Mean procedure time was 39.4 minutes.
There were no adverse events during or immediately following the procedure.
Table 2
Endoscopic ultrasound (EUS) characteristics.
|
Diagnosis
|
|
|
Benign, n (%)
|
2 (7.7)
|
|
Malignant, n (%)
|
24 (92.3)
|
|
Diagnostic accuracy
|
|
|
FNA, n (%)
|
24 (92.3)
|
|
Core biopsy, n (%)
|
22 (84.6)
|
|
EUS-FNA test characteristics
|
|
|
Sensitivity, %
|
83
|
|
Specificity, %
|
100
|
|
EUS core biopsy test characteristics
|
|
|
Sensitivity, %
|
91.7
|
|
Specificity, %
|
100
|
|
Number of passes for FNA, n
|
|
|
Mean
|
3.2
|
|
Range, min, max
|
2, 7
|
|
Procedure time, min
|
|
|
Mean
|
39.4
|
|
Range, min, max
|
15, 80
|
EUS, endoscopic ultrasound; FNA, fine-needle aspiration.
The final diagnosis was malignant in 92.3 % of the cases and benign in 7.7 % of the
cases. Diagnostic accuracy was 84.6 % (95 %CI: 66.4 – 93.8 %) for EUS core biopsy
and 92.3 % (95 %CI: 75.5 – 97.8 %) for EUS-FNA. The difference in accuracy between
the two approaches was not statistically significant (P = 0.14). The kappa statistic, which was calculated to measure the agreement in yield
between EUS-FNA and EUS-guided core biopsy, was 0.62 (95 %CI 0.33 – 0.91). The sensitivity
and specificity for EUS-FNA were 83 % and 100 %, respectively. The sensitivity and
specificity for EUS core biopsy were 91.7 % and 100 %, respectively.
Discussion
Endoscopic ultrasound-guided FNA with on-site cytopathology has become the gold standard
in the evaluation of gastrointestinal and peri-intestinal mass lesions. This practice
is not only cost-effective [13] but immediate review of FNA by an on-site cytopathologist has been shown to increase
diagnostic yield by as much as 18 – 26 % [6]
[7]. Unfortunately, this practice has not been universally embraced, most likely due
to cost as well as lack of the necessary expertise and personnel. As such, there have
been more and more studies published which examine alternatives to the on-site cytopathologist.
One such alternative is the use of EUS-guided core biopsy [14]. Based on the results of our study, the diagnostic yield of two passes with a 22-gauge
core biopsy needle is comparable to EUS-FNA with on-site cytopathology. The yield
of EUS-FNA with a bedside cytopathologist in our study, irrespective of EUS-guided
core biopsy, is similar to the yields published in other studies, which range from
78 % to 89 % [15]
[16]
[17]. For institutions which cannot afford or do not have access to an on-site cytopathologist,
performing two core biopsies may improve yields and prevent unnecessarily repeating
the procedure. As far as we can tell, this is the only study which compares EUS core
biopsy to our current gold standard of EUS-FNA in the same patients with on-site cytopathology
review.
At this time, most of the published studies compare the diagnostic yield of EUS core
biopsy using a 19-gauge needle and EUS-FNA without bedside cytopathology [16]
[17]
[18]
[19]. While diagnostic accuracy tends to vary depending on the site, there is generally
no significant difference between the two modalities [17]
[18]
[19]. While the 19-gauge Trucut core biopsy needle appears to operate well at certain
sites such as the esophagus and portions of the stomach, it was more difficult to
use in the antrum/fundus of the stomach as well as the duodenal bulb [20]. Even with the new, European-designed 19-gauge fine needle biopsy device, which
was designed to overcome the limitations of obtaining transduodenal samples from the
pancreatic head, this process continues to be technically difficult [21]. As a result, other types of FNA and core biopsy needles are being developed and
compared [22].
Bang et al. published a randomized trial which compared the diagnostic yield of the
22-gauge FNA needle and a new 22-gauge biopsy needle for EUS-guided sampling of solid
pancreatic masses [23]. With the new 22-gauge biopsy needle, they were able to obtain transduodenal biopsies
without difficulty, thus overcoming the limitations of the 19-gauge core biopsy needles.
Interestingly, the new 22-gauge core biopsy needle was capable of obtaining cytology
and histology specimens. Their study concluded that the diagnostic yield of the new
22-gauge biopsy needle is comparable to the 22-gauge FNA needle. The authors commented
that the yield of histologic core tissue was unsatisfactory with the biopsy needle,
but there was no statistically significant difference in the number of passes for
diagnosis or number of cases where there was a failure to achieve the diagnosis between
the two diagnostic modalities. Another study examined 62 patients with solid pancreatic
lesions which were sampled by EUS-guided 22-gauge core biopsy needle and 25-gauge
FNA needle at the same endoscopic session [24]. There was no difference in adequacy of the specimens obtained through FNA and core
biopsy needles. Additionally there was a significant agreement between EUS-FNA and
core biopsy (88.5 % for positive agreement and 62.5 % for negative agreement).
In our study, EUS-guided core biopsies were obtained before EUS-FNA of the same lesion.
As previously discussed, one of the inherent limitations of FNA is the lack of architecture
which can be important in making certain diagnoses. This decision was made because
we did not want the FNA to disrupt the underlying architecture and diminish the yield
of core biopsy although this is only a theoretical risk. Additionally, we did not
alternate the two modalities in an effort to keep more variables constant. It would
be interesting to see whether results would be similar if EUS-FNA preceded the EUS-guided
core biopsies.
There are several advantages to our study. It includes a wide spectrum of disease
which is not limited to solid pancreatic mass lesions. It compares the yield of EUS-FNA
and EUS core biopsy performed on the same lesion during a single endoscopic session.
Studies which compare the yield of these two modalities generally use a criterion-standard
reference method. While patients with suspected malignancies based on either of these
diagnostic tests generally undergo a subsequent surgical resection whereby the final
diagnosis may be confirmed, those with benign cytology or histology are followed clinically.
The final diagnosis in these “benign” cases is generally determined after a certain
period of time based on a patient’s clinical course and/or subsequent studies which
may include imaging or repeat endoscopy with or without sampling. The criterion-standard
reference method is used in our study as well, but mass lesions were sampled simultaneously
using EUS-FNA and core biopsies. In this manner, patients serve as their own control.
Also, all core biopsy specimens were reviewed by a single pathologist who was blinded
to the results of the preceding FNA specimens.
Despite performing EUS-guided core biopsies followed by multiple passes for FNA during
the same endoscopic session, the procedure was performed safely and efficiently. There
were no adverse events during or immediately following the procedure. Mean procedure
time was 39.4 minutes. EUS-FNA has been accepted to be a safe intervention with a
low post-procedural adverse events rate [25]. EUS-guided core biopsy (using the 19-gauge Trucut needle) has also been shown to
be safe, with an adverse events rate of approximately 2 % [26]. Although more studies are needed, the EUS-guided core biopsy may eventually supplant
EUS-FNA with on-site cytopathology as the gold standard.
Limitations
Much like other published studies examining the diagnostic yield of EUS-FNA and EUS
core biopsy, the primary limitation of this study is the small size of the study population.
Additionally, we examined a heterogenous population of lesions in this study. Also,
some patients referred for biopsy were not included because core biopsy could not
be performed. As such, results are biased towards patients in whom core biopsy was
technically feasible. While all EUS procedures were performed by a single, experienced
endoscopist in our study which allows for standardization, diagnostic yield may vary
at other institutions when endoscopy is performed by multiple physicians or less experienced
physicians. This study was conducted at a single tertiary referral center, and as
such, there was a disproportionate amount of malignant lesions. Also, a cost analysis
was not performed in this study but it would be interesting to see whether EUS core
biopsies are cost effective when compared with EUS-FNA + on-site cytopathologist.
Lastly, our study was not designed or powered to show significant differences between
EUS-FNA and core biopsies but to compare the accuracy of both approaches.
Conclusion
Based on our study, the diagnostic yield of two passes with a 22-gauge core biopsy
needle may be comparable to the current gold standard of FNA with a bedside cytopathologist
when sampling gastrointestinal lesions. Large, prospective, randomized studies are
still needed to further compare these two modalities. Eventually, an approach with
two core biopsies could represent a time efficient and widely available alternative
to FNA with a bedside cytopathologist.
