Diagnostic accuracy of fine-needle aspiration of solid pancreatic lesions guided by endoscopic ultrasound elastography

Patients

From a prospectively collected database, we reviewed data for 54 consecutive patients out of 83 subjects with solid pancreatic lesions, referred to our institution between January 2014 and June 2015. All patients had been diagnosed with pancreatic masses by means of transabdominal ultrasonography, computed tomography scan or magnetic resonance imaging. The Institutional Review Board approved this retrospective report and written informed consent was obtained from all the patients.

The following exclusion criteria were used: age under 18 years, cystic pancreatic lesions, lesions <1 cm, history of previous gastrectomy, severe coagulopathy or current anticoagulant/antiaggregant therapy that could not be suspended, inability or refusal to provide informed consent. All procedures were performed by a board-certified gastroenterologist (NM), who had performed more than 600 EUS-FNA and 150 RTE-EUS before the study period.

Technical procedure

The technique used for diagnostic and interventional EUS at our center has been described elsewhere [11,12]. Briefly, with the patient under sedation with propofol, EUS was conducted using a Pentax FG-36UA ultrasound endoscope (Pentax Europe, Ltd., Hamburg, Germany) with a curved-array transducer. After introduction to the stomach, the endoscopic ultrasound probe was placed in contact with the gastric wall and the aorta was identified in an elongated cross-section—a finding confirmed by color Doppler imaging. The scope was then slowly advanced up to the duodenum when necessary. A 25G needle (EchoTip Ultra, Cook Medical LLC, Bloomington, Indiana, USA), with a central stylet to protect the needle’s aspiration channel, was introduced through the endoscope’s working channel. A transduodenal approach was used for lesions of the pancreatic head and uncinate process, whereas the trans-gastric approach was used for lesions of the body or tail.

RTE assessment of pancreatic masses was performed using a Hitachi Hi Vision Ascendus device (Hitachi medical systems, Tokyo, Japan). All areas that appeared suspicious on elastography (i.e., those appearing in a dark blue color as a consequence of the higher cellularity of tumoral tissue) were recorded and stored in our database. Both a qualitative assessment, based on a red-green-blue color map, and a semi-quantitative approach that provided a numeric value expressed as strain ratio [13] were undertaken.

The needle was inserted into the most suspicious part (“dark blue”) of the lesion and immediately after the procedure the stylet was removed. More than 10 to-and-fro movements were made within the lesion and aspiration was obtained with a 10 cm3 suction syringe applied to the hub of the FNA device. Two passes were performed in most patients. At the end of the procedure, the needle was retracted and the samples were fixed in 95% ethanol solution. After being grossly checked for adequacy, samples were prepared for cytological examination with Papanicolaou staining. Patients were continuously monitored during the procedure by a board-certified anesthesiologist with an automated noninvasive blood pressure device, electrocardiogram tracing and pulse oximetry. The procedure adopted in the current study is depicted in Fig. 1.

Figure 1 An endoscopic ultrasonography image describing real-time elastography-targeted fine-needle aspiration of a solid lesion located in the pancreatic head. The needle was inserted into the most suspicious part (“dark blue”) of the lesion. The final diagnosis identified ductal adenocarcinoma

Cytology

The degree of cellularity and blood contamination of specimens was graded as five levels (0: none; 1: minimal; 2: mild; 3: moderate; 4: many) and cellularity was then divided into two groups: “high cellularity” indicated a cellular level of 3 or 4 and “low cellularity” indicated a level of 0, 1 or 2 [14].

The reference standard for classification was surgery or death from pancreatic cancer in those subjects unsuitable for surgery. In particular, if after a follow up of 12-18 months there was no sign of disease progression, or if disease regression was registered, the lesion was classified as inflammatory [15].

Lesions diagnosed as malignant by cytopathology of the EUS-FNA sample and were finally confirmed by surgery or clinical course were considered to be true positives; likewise, benign aspirates finally diagnosed as benign were considered to be true negatives. On the other hand, aspirates, apparently benign on cytopathological examination but finally diagnosed as malignant, were considered to be false negatives. Non-diagnostic or inconclusive samples were registered as such in the database and for analytical purposes were classified as false negatives when diagnostic accuracy was computed.

Follow up

All clinical and safety outcomes, as well as imaging follow up, were assessed by clinicians blinded to the EUS technique adopted. Adverse event rates were evaluated during the procedure, before discharge, and at 1 and 3 days by means of telephone calls. After the first month, all the clinical visits, which included physical examination, laboratory analyses (including tumor markers) and imaging evaluation, were performed on an outpatient basis at 50 days after the procedure and every 3 months thereafter.

Statistical analysis

Categorical variables were expressed as frequencies and percentages and continuous variables as medians and interquartile ranges (IQRs). Sensitivity, specificity, diagnostic accuracy, positive and negative predictive values were calculated. Sensitivity was computed as the proportion of positives correctly identified by the test among the prevalence of malignancy in the study cohort. Specificity was calculated as the proportion of negatives correctly identified as such among the patients not affected by the tumor in the study cohort. Positive likelihood ratio was defined as sensitivity/false-positive rate and negative likelihood ratio as false-negative rate/specificity. The analysis was performed using R Statistical Software (Foundation for Statistical Computing, Vienna, Austria) and significance was established at the 0.05 level (two-sided).

Patients

Table 1 summarizes the main characteristics of the study population. The median age was 68 years (IQR: 59-73) and 64.8% of patients were male. The majority of lesions were located in the pancreatic head or uncinate (62.9%) and the median size was 35 mm (IQR: 25-43). About one fifth of the whole population (20.3%) finally underwent surgery and the diagnosis of adenocarcinoma was confirmed from a surgical specimen or by the patient’s clinical course in 85.1% of cases.

Table 1 Baseline characteristics of the study population

RTE findings

The quantitative elastography score was represented by the strain ratio method, i.e., as the ratio between a surrounding non-mass area (area B) and the mass or region of interest (area A) [13]. Based on the results of the receiver operating characteristic curve used for analysis (Fig. 2), the cutoff level of strain ratio able to achieve the maximal accuracy (expressed as area under the curve) was 4.21, which provided a sensitivity of 86.9%, a specificity of 75%, and a diagnostic accuracy of 85.1% (Table 2). When the cutoff value of 6.04 (proposed in the landmark study by Iglesias-Garcia et al [13]) was adopted, specificity increased significantly to 87.5% but sensitivity dropped to 80.4%, resulting in a slight decrease in overall accuracy (81.4%) (Table 2). Pancreatic ductal adenocarcinoma appeared as homogeneous hard (blue) tissue in 20% of cases, while we found areas with different strain (mainly due to necrosis) that allowed targeting the hardest area specifically in 80% of lesions.

Table 2 Elastography findings

Figure 2 Receiver operating characteristic (ROC) curve analysis aimed at identifying the best cutoff point for elastography strain ratio. ROC curve analysis identified the value 4.21 as the most sensitive and specific cutoff point for the elastography strain ratio

Diagnostic accuracy of RTE-targeted EUS-FNA

All diagnostic outcomes are reported in Table 3. No needle malfunctions nor mechanical failures occurred and technical success was 100%. A cytological diagnosis positive for malignancy was found in 43 cases (79.6%), while 10 samples were negative (18.6%) and in one single case the cytological sample was rated as inadequate (1.8%).

Table 3 Diagnostic performance of the procedure

Diagnostic accuracy, sensitivity, and specificity were 94.4%, 93.4% and 100%, respectively. Since no false positives were observed, positive predictive value was 100% and positive likelihood ratio could not be computed (false-positive rate, the denominator of the proportion, was 0). On the other hand, 3 false negatives were registered; hence, the negative predictive value was 72.7% and the negative likelihood ratio was 6.5 (Table 3).

Cytological findings and adverse events

According to commonly accepted scores published elsewhere [14,15], overall cellularity was rated as average (2.5, IQR: 1-4), as were blood contamination (2, 0-3) and neoplastic cellularity (2, 1-4). Contamination by cellular debris or amorphous material did not represent a real issue, since a low rate for this specific item was given by the cytopathologist (1, 0-2) (Table 4).

Table 4 Cytological findings

No severe adverse events were registered. Five patients experienced a slight increase in pancreatic enzyme levels (9.2%) and 3 complained about mild abdominal pain (5.5%). Mild post-procedural bleeding was observed in 2 cases (3.7%). All of these complications were managed conservatively.